Greening Through Recognition: Unveiling the Mechanisms of China’s High-Tech Enterprise Identification Policy on Sustainable Innovation

Abstract

This study examines whether China’s high-tech enterprise identification policy promotes corporate sustainable innovation. Using panel data from Chinese listed firms on the Shanghai and Shenzhen stock exchanges between 2008 and 2022, we adopt a time-varying difference-in-differences (DID) model to evaluate the policy’s effectiveness and explore its underlying mechanisms. The results reveal that this certification policy significantly facilitates green innovation, and the findings remain robust across various checks, including alternative measurements, placebo tests, propensity score matching DID (PSM-DID), and the exclusion of digital transformation trend and confounding macro-level policies. Mechanism analysis shows that the policy influences green innovation by alleviating financing constraints, increasing access to government subsidies, facilitating the agglomeration of scientific and technological talent, and encouraging greater R&D investment. Heterogeneity analysis further indicates that the policy effect is more pronounced among non-state-owned enterprises, small-scale firms, capital-intensive businesses, those located in high-institutional-quality regions, and firms in China’s eastern provinces. Moreover, the positive impact is strongest for growth-stage firms. The policy has also been found to improve green innovation efficiency. These findings offer empirical insights for optimizing selective industrial policies to enhance sustainable innovation and support China’s dual-carbon goals.

1. Introduction

Corporate green innovation, also known as eco-innovation, plays a critical role in supporting the carbon neutrality target and fostering sustainable development. Green innovation can enhance firms’ financial performance [1], expand market share [2], and increase overall business value [3], thereby strengthening competitiveness [4]. However, due to the high costs, long R&D cycles, and uncertain returns [5], green innovation poses substantial risks and typically generates positive externalities that discourage firms from initiating such efforts without policy incentives [6].

The Chinese government has implemented various policy tools, such as green credit [7] and low-carbon pilot city initiatives [8], to promote green innovation. Among these, selective industrial policies have emerged as key instruments for supporting strategic emerging industries like artificial intelligence, biotechnology, and clean energy [9]. Compared with general industrial policies, selective policies are more targeted and provide greater incentives to firms with strategic potential. China’s high-tech enterprise identification policy (HTEIP) serves as a representative example of such policies. The “Measures for the Administration of the Identification of High-Tech Enterprises” (MAIHTE) was issued by the Chinese government on 14 April 2008, and these measures include the following certification criteria: the business should focus on key industrial technology fields supported by the state, and the main business operations should be closely centered on their R&D and technology commercialization activities. Enterprises that obtain certification enjoy various policy benefits, including tax incentives, accelerated depreciation of fixed assets, and additional R&D deductions. Existing studies have proved the promotion effects of this policy on firms’ R&D intensity, productivity, and technological innovation [10,11].

In terms of the factors that affect green innovation, previous studies have comprehensively investigated those from both the internal and external dimensions of enterprises. Internally, financing support plays a critical role. Green innovation is a long-term, high-risk process requiring stable and substantial financial input. Studies show that alleviating financing constraints boosts firms’ eco-innovation capacity [12]. For example, digital finance can ease funding limitations and promote green innovation [13], and green credit policy could improve innovation performance by easing firms’ financial difficulties [14]. Other internal factors include corporate social responsibility [15], green dynamic capabilities [16], and CEO characteristics such as overseas experience [17], political connections [18], and gender [19]. Moreover, knowledge management capacity significantly enhances firms’ green innovation performance [20].

Externally, moderate environmental regulations and government support are major drivers. Moderate regulation can stimulate green innovation by exerting pressure and offering incentives [21,22,23]. Public attention also plays a positive role in shaping eco-innovation activities [24]. Government subsidies, including those for R&D, environmental protection, and talent, directly promote green innovation [25,26]. Furthermore, the rise of the digital economy creates favorable conditions by optimizing industrial structures and expanding market potential [27], thereby promoting firms’ green innovation [28,29].

Existing research on the HTEIP has confirmed its effectiveness in stimulating innovation [30], though concerns about R&D manipulation also exist [31]. Additionally, financial constraints and higher supplier concentration could weaken the innovation effects of the policy [32]. However, empirical studies that directly examine its impact on green innovation, especially regarding the specific mechanisms, remain scarce.

In light of this research gap, this study investigates the effect of China’s HTEIP on green innovation by utilizing panel data from Chinese listed firms from 2008 to 2022. Employing a time-varying difference-in-differences (DID) approach and green patent data at the firm level, we assess not only whether the policy promotes eco-innovation, but also through which pathways such effects occur.

Our marginal contributions are as follows. First, our study offers a micro-level evaluation of the policy’s green innovation incentives, both theoretically and empirically. Second, we examine the heterogeneity of policy effects across various perspectives such as ownership, firm size, industry factor intensity, institutional environment, and geographic region. Third, we explore multiple mechanisms, including financing constraints, government subsidies, talent agglomeration, and R&D investment. Finally, further analysis of the policy’s incentivizing heterogeneity from the corporate life cycle and green innovation efficiency perspectives is conducted. These insights enrich the understanding of how selective industrial policies can facilitate corporate green transformation and provide evidence-based policy implications for refining such initiatives.

2. Theoretical Analysis and Hypotheses

2.1. The Mediating Role of Alleviating Financing Constraints

The high-tech enterprise identification policy can ease both the external and internal financing constraints of enterprises through external signaling effects and increased internal cash flows. First, the identification policy could reduce information asymmetry in the financial markets through an external signaling effect, which helps alleviate financing constraints. Identified enterprises can signal their good credit, strong technological innovation capabilities, and promising development prospects in the market. These positive signals can help enterprises obtain bank loans or credit support from other financial institutions more easily and attract more venture capital, thus alleviating financing constraints. Second, adjusting the tax loss carry-forward period for high-tech enterprises helps ease financial pressure. According to the latest regulations issued by the MOF and SAT, the period of tax loss carry-forward can be extended from the previous 5 years to 10 years, which means that the identified high-tech enterprises will bear less financial pressure brought by short-term losses, allowing for more flexible fund allocation. Additionally, firms that achieve high-tech identification enjoy a preferential tax rate of 15–25%, which helps reduce the financial burden and cash outlay of companies [33,34].

Financing expansion helps encourage enterprise green innovation. Eco-innovation is a strategic transition that is typically long-term and high-risk, which usually requires substantial financial support. The pressure to finance may reduce firms’ motivation to carry out eco-innovation. Alleviating financing constraints helps enterprises increase inputs in green technology R&D. Based on the introduction in the previous section and the above analysis, this study proposes the following hypotheses:

Hypothesis 1 (H1).

The high-tech enterprise identification policy promotes firms’ green innovation. After attaining high-technology enterprise status, a company’s degree of green innovation increases over time.

Hypothesis 2a (H2a).

The high-tech enterprise identification policy promotes firms’ green innovation by alleviating financing constraints.

2.2. The Mediating Role of Government Subsidy Effect

The identification policy enables firms to obtain more government subsidies, which promotes enterprise green innovation. The identified enterprises can benefit from various national, provincial, and municipal preferential policies, including access to research funds, financial grants, and government subsidies, such as innovation subsidies, project loan interest subsidies, environmental protection subsidies, and land concessions. A positive relationship has been found between government subsidies and firms’ green innovation [26,35]. Based on these inferences, we put out the following hypothesis:

Hypothesis 2b (H2b).

The high-tech enterprise identification policy promotes enterprise green innovation through government subsidy effects.

2.3. The Mediating Role of Talent Agglomeration Effect

Enterprises that achieve high-tech enterprise identification have the advantage of clustering scientific talent. First, the high-tech enterprise identification policy promotes scientific and technological (S&T) talent agglomeration through resource advantages [36], attracting potential and ambitious experts. Second, according to the standards of the policy, the percentage of employees involved in technological R&D activities must be at least 10% of the total number of employees in that year. This requirement compels enterprises to focus on recruiting and cultivating S&T research personnel to meet certification standards. Finally, this policy provides substantial financial support to enterprises, significantly enhancing employee salaries and benefits. Thus, it attracts highly qualified research personnel and reinforces talent agglomeration.

The agglomeration of highly qualified human resources enhances corporate green innovation performance [37]. First, talent aggregation facilitates knowledge sharing, knowledge flow, and collaboration within an enterprise, driving the innovation of green technologies. Second, highly qualified talent possesses greater flexibility and adaptability, enabling firms to respond quickly to environmental standards and market demand, thereby helping firms establish core competitiveness in the green transition. Based on this analysis, we put forward the following:

Hypothesis 2c (H2c).

The high-tech enterprise identification policy promotes firms’ green innovation through the talent agglomeration effect.

2.4. The Mediating Role of R&D Investment

Enterprise green innovation R&D investments often face high risks, uncertain returns, and significant adjustment costs [38]. Additionally, the lack of liquidity in green R&D investment and the difficulty in obtaining external financial support [39] may reduce the firm’s willingness to invest in R&D. The high-tech enterprise identification policy can increase corporate R&D investments through policy incentives and self-monitoring. First, the identification policy provides clear guidance and strong incentives for enterprises to increase their R&D inputs by establishing explicit quantitative standards. Specifically, the policy requires that an enterprise’s total R&D expenditures threshold be based on the firm’s size. For example, it must be no less than 5% of its total sales revenue in the last three accounting years for a smaller enterprise. Second, the certification assessments clarified the handling measures for enterprises that do not meet the certification conditions, including the cancellation of their high-tech enterprise certification and the recovery of the tax concessions they have enjoyed. This review mechanism stimulates enterprises’ intrinsic motivation, promotes self-supervision, and encourages them to continually increase their R&D inputs. Then, increased R&D inputs boost green innovation activities [40]. R&D investment supports the development of green products, efficient production processes, and eco-friendly materials, improves resource efficiency, and reduces resource waste. Therefore, we propose the following hypothesis:

Hypothesis 2d (H2d).

The high-tech enterprise identification policy promotes firms’ green innovation through increasing R&D investment.

3. Methodology

3.1. Data Source and Sample Selection

This study used data from Chinese A-share listed companies on the Shanghai and Shenzhen stock exchanges from 2008 to 2022. To enhance the representativeness of the sample, the original sample was adjusted as follows: (1) observations with severe data omissions or obvious financial anomalies were excluded; (2) ST and delisted firms were excluded; (3) companies in financial and real estate industries were excluded; and (4) continuous variables were winsorized on both tails at 1% level to ensure robustness. Corporate financial information and high-tech enterprise identification data were obtained from the CSMAR database. Data on corporate green innovation were sourced from the China Research Data Services (CNRDS) database. After processing, the final sample comprised 4246 companies, of which 2751 with high-tech certification and 1495 without. And the sample used for empirical analysis included 35,141 firm-year observations.

3.2. Models

To examine the impact of the HTEIP on firms’ green innovation activities, following the study of [11], this paper constructed the following baseline multi-period DID model:

$${GreenInnovation}_{it}={\alpha }_0 +{\alpha }_1{Hightech}_{it}+{\lambda }^{\prime }{Controls}_{it}+{Firm}_i+{Year}_t +{\epsilon }_{it}$$

(1)

In Formula (1), i denotes the firm and t represents the year. GreenInnovation is the measure of enterprise green innovation, calculated by the number of green invention patents obtained by a company. Hightech_it is the core explanatory variable, indicating whether firm i obtained high-tech enterprise identification in year t; Controls represents a range of control variables at the firm-level and ${\lambda }^{\prime }$ represents the vector of estimated coefficients associated with the firm-level control variables; Firm and Year denote firm- and year-fixed effects, respectively. ε is the random error term.

Table 1. Descriptive statistics.

Variable	Observation	Mean	Std.Dev	Max	Min
GreenInnovation	35,141	0.138	0.433	2.565	0.000
Age	35,138	2.811	0.380	3.497	1.609
Size	34,069	7.655	1.236	11.190	4.970
ln_Board	35,141	2.296	0.461	2.944	0.000
Top1_ratio	35,141	1.379	1.630	4.615	0.000
Leverage	35,141	0.328	0.142	0.624	0.049
ROA	35,131	0.042	0.061	0.197	−0.235
Mab_ratio	35,137	0.439	0.378	1.098	0.000
ROE	35,134	0.037	0.053	0.186	−0.182
Turnover	35,141	0.378	0.269	1.197	0.000
Growth	35,141	0.156	0.242	1.174	−0.298

presents the results of descriptive statistics. Detailed definitions of all variables are shown in Appendix A.

3.3. Variables

(1)

Enterprise green innovation (GreenInnovation). Green patents are primarily distinguished into three categories: green invention patents, green utility models, and green design patents, with innovation difficulty progressively decreasing across these three categories [41,42]. Comparatively, green invention patents could better reflect firms’ green innovation capabilities. Therefore, based on data availability and representativeness, this study used the logarithm of one plus the quantity of authorized green invention patents as the measure of eco-innovation behavior.

(2)

High-tech enterprise identification policy (Hightech_it). The core explanatory variable, Hightech_it, is an interaction term of Post and Treat. Post is a time dummy variable indicating the implementation period of the identification policy. It equals 1 for the year when a firm attained high-tech status and for all subsequent years, and 0 otherwise. Treat distinguishes between the control and treatment groups. If a firm obtained high-tech identification during the sample period, it is considered part of the treatment group, with Treat equal to 1; otherwise, it equals 0.

(3)

Control variables. Drawing on the studies [43,44,45], a set of controls are included: (1) Basic enterprise characteristics: firm’s age (Age), firm’s size (Size), logarithm of board size (ln_Board), the largest shareholder’s holding ratio (Top1_ratio), liability/asset ratio (Leverage), return on asset (ROA), the proportion of the main business (Mab_ratio), return on equity (ROE), total asset turnover (Turnover), total asset growth rate (Growth).

4. Empirical Results and Discussion

4.1. Benchmark Regression

Table 2. Results of baseline model test.

	(1)	(2)	(3)	(4)
Hightech	0.0681 ***	0.0192 ***	0.0188 ***	0.0186 ***
	(14.7879)	(2.8358)	(2.7724)	(2.7271)
Age		0.0753 ***	0.0748 ***	0.0604 **
		(2.6647)	(2.6439)	(2.0892)
Size		0.0390 ***	0.0390 ***	0.0368 ***
		(8.2407)	(8.2482)	(7.5359)
ln_Board			0.0049	0.0033
			(0.4631)	(0.3141)
Top1_ratio			0.0043 ***	0.0044 ***
			(3.0571)	(3.0932)
Leverage				0.0428 *
				(1.7015)
ROA				−0.0377
				(−0.7213)
Mab_ratio				0.0053
				(0.7286)
ROE				−0.0440
				(−0.6695)
Turnover				0.0010
				(0.0857)
Growth				−0.0167 **
				(−2.1147)
Constant	0.1055 ***	−0.3766 ***	−0.3928 ***	−0.3425 ***
	(33.1940)	(−4.5661)	(−4.5449)	(−3.8727)
Firm	No	Yes	Yes	Yes
Year	No	Yes	Yes	Yes
Observations	35,141	33,573	33,573	33,561
R-squared	0.0062	0.6271	0.6273	0.6274

Note: t-values in parentheses. * p < 0.1, ** p < 0.05, and *** p < 0.01. The same notation applies to the following tables.

reports the results of the benchmark regression model. Column (1) gives the estimated results considering only the core independent variable Hightech without accounting for firm- and year-fixed effects. The results show that the coefficient of Hightech is 0.0681, which is significant at the 1% level. Column (2) further adds firm’s size and firm’s age controls; the regression coefficient of Hightech is 0.0192 at the 1% significance level. Column (3) includes the controls Board and Top1_ratio; the coefficient of Hightech is 0.0188 at the 1% level. Finally, when all the control variables are considered in the last column, the coefficient of Hightech is 0.0186 and remains statistically significant. Therefore, the financing, subsidy, talent agglomeration, and R&D investment effects of the identification policy significantly promote enterprise green innovation, thus verifying H1.

4.2. Parallel Trend Test

This study used the DID model to estimate the impact of the HTEIP on green innovation. The underlying assumption is that high-tech and non-high-tech companies have the same time trends before policy implementation. It implies that, without the policy intervention, the eco-innovation performance of high-tech enterprises should follow the same development trend as non-high-tech enterprises. Following Beck et al. [46], we established the dynamic regression model to test the parallel trend assumption as follows:

$${ GreenInnovation }_{it}={\alpha }_0+\sum _{k=-2}^{k=-14}{\beta }_{k}pre+{\alpha }_{1}current+\sum _{m=1}^{m=14}{\alpha }_{m}last+\phi {Controls}_{it}+{Firm}_i+{Year}_t +{\epsilon }_{it}$$

(2)

In the dynamic regression model, k = t − year0, where t represents the sample year and year0 is the year when the enterprise first obtained high-tech enterprise identification. If the enterprise was not identified, year0 is set to 0. This study used the year when the policy was implemented (denoted as period “0”) as the base period. In Formula (2), the second term represents the dynamic effects of the high-tech policy before its implementation; the third term denotes the effect of the policy in the current year of implementation; the fourth term captures the lagged effects of the policy after implementation.

As shown in Figure 1, the coefficients of the policy impact are close to 0 and not significant before the policy shock, thus satisfying the parallel trend. The coefficients of the policy impact gradually increase and become significant after the policy intervention. Furthermore, the upward trend of post-intervention indicates that this policy has a positive effect on companies’ green innovation.

4.3. Robustness Checks

4.3.1. Counterfactual Tests

In addition to the high-tech enterprise identification policy, enterprise green innovation activities may also be influenced by related policies and economic development strategies. Therefore, this study conducted a series of counterfactual tests. Specifically, the timing of high-tech enterprise certification was advanced by one year (F1.Hightech), two years (F2.Hightech), and three years (F3.Hightech) to test the robustness.

Table 3. Results of counterfactual test.

	(1)	(2)	(3)
F1.Hightech	0.0004
	(0.0447)
F2.Hightech		0.0069
		(0.7089)
F3.Hightech			0.0134
			(1.2932)
Constant	−0.2897 ***	−0.2429 **	−0.3202 ***
	(−3.0535)	(−2.4029)	(−2.9650)
Controls	Yes	Yes	Yes
Firm	Yes	Yes	Yes
Year	Yes	Yes	Yes
Observations	28,477	24,276	20,999
R-squared	0.6268	0.6293	0.6324

Note: t-values in parentheses. ** p < 0.05, and *** p < 0.01.

presents the results. In these three counterfactual settings, if the dummy variables for the advance period do not significantly affect corporate green innovation, the conclusions drawn from the baseline model are credible. Conversely, if the dummy variables show significant effects, this suggests that green innovation may benefit from other policies, and the conclusions of the baseline model would be unreliable. In Table 3, the coefficients of F1.Hightech, F2.Hightech, and F3.Hightech are insignificant. This indicates that advancing the identification policy by one, two, or three years has no significant impact. This indicates that the benchmark results are not affected by other policies; therefore, the conclusions obtained from the benchmark model are credible.

4.3.2. The Interference of Digital Transformation and Other Macro Policies

According to the theoretical analysis, the high-tech enterprise identification policy can alter firms’ financing behaviors, thereby affecting green innovation activities. Simultaneously, firms’ digital transformation can affect enterprise eco-innovation by influencing enterprise financing patterns [47]. To exclude the possible impacts of digital transformation on the estimation results, this study included the digital transformation variable (Digital) in the model and re-tested the results. By examining the frequency of digital keywords in listed firms’ annual reports, this study conducted an exhaustive text analysis. It quantitatively assessed the level of digital transformation by calculating the ratio of the frequency of digital keywords to the total text volume of annual reports. To enhance precision, the indicators were expressed in percentage form. A higher value of Digital means a higher level of digital transformation within the enterprise. The regression results are reported in

Table 4. Excluding the consequences of enterprise digital transformation and other macro policies.

	(1)	(2)	(3)	(4)	(5)
Hightech	0.0200 ***	0.0186 ***	0.0201 ***	0.0185 ***	0.0201 ***
	(2.8523)	(2.7329)	(2.8690)	(2.7167)	(2.8682)
Digital	3.1984 ***		3.2034 ***		3.2056 ***
	(7.8238)		(7.8308)		(7.8361)
Smart		0.0041	0.0044		0.0048
		(0.9093)	(0.9622)		(1.0285)
BroadBand				−0.0009	−0.0018
				(−0.1990)	(−0.3925)
Constant	−0.3452 ***	−0.3449 ***	−0.3477 ***	−0.3425 ***	−0.3471 ***
	(−3.8553)	(−3.9032)	(−3.8835)	(−3.8752)	(−3.8762)
Controls	Yes	Yes	Yes	Controls	Yes
Firm	Yes	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes	Yes
Observations	32,629	33,570	32,617	33,570	32,617
R-squared	0.6190	0.6274	0.6190	0.6274	0.6190

Note: t-values in parentheses. *** p < 0.01.

. It can be seen in Column (1) that the coefficient of Hightech is still significantly positive after excluding the effect of digital transformation.

Second, Chinese smart city policies [48] and the “Broadband China” pilot policy can also promote green technological innovation. Therefore, to control for the potential impact of these policies, we add two policy dummy variables, smart city pilot indicator (Smart) and “Broadband China” pilot city (BroadBand), to the model, respectively. The results in Table 4 show that Hightech remains significant at the 1% level when we consider both digital transformation and other macro policies.

4.3.3. The Sample Selection Bias Issues: PSM-DID Estimation

The sample of firms identified as high-tech enterprises may not have been randomly selected, which could result in a potential sample selection bias. This study adopted the PSM-DID approach and conducted 1:2 nearest neighbor matching, 1:3 nearest neighbor matching, and kernel matching to address this issue. First, the control variables were used to perform propensity score matching, as mentioned in the previous section. Each treatment group observation was matched with the control group observations with the closest propensity scores. Additionally, a balance check was conducted to verify the matching algorithm. We plotted the kernel density of the propensity scores, and the results are shown in Figure 2 and Figure 3. The figures indicate that the distribution of propensity scores between the two groups was close, which is consistent with the matching principle.

Table 5. Results of PSM-DID.

	(1)	(2)	(3)
	1:2 Nearest Neighbor	1:3 Nearest Neighbor	Kernel Matching
Hightech	0.0192 *	0.0186 **	0.0160 **
	(1.8677)	(2.1449)	(2.3724)
Constant	−0.2481 **	−0.3643 ***	−0.2861 ***
	(−2.1095)	(−3.5409)	(−3.0505)
Controls	Yes	Yes	Yes
Firm	Yes	Yes	Yes
Year	Yes	Yes	Yes
Observations	22,570	26,887	29,489
R-squared	0.6538	0.6460	0.6131

Note: t-values in parentheses. * p < 0.1, ** p < 0.05, and *** p < 0.01.

shows the estimation results of the matched samples. The coefficients of Hightech are 0.0192, 0.0186, and 0.0160, which are statistically significant at the 10% and 5% significance levels. Therefore, the findings of the estimation after PSM processing are in line with those of baseline regression.

4.3.4. Placebo Text

To minimize the impacts of other factors on the incentive effects of the identification policy, this study conducted a placebo test. Referring to Li et al. [49], a random year within the sample period was selected as the point at which firms obtain high-tech enterprise identification, and the estimation was repeated. After 500 simulation experiments, we observed the phenomenon shown in Figure 4, where the effects of the policy variable reappeared, with the coefficients remaining around 0, approximately obeying a normal distribution, and the p-values being above 10%. This clearly shows that the impact of this policy on enterprise green innovation is not coincidental, further demonstrating the robustness of this study’s conclusions.

4.4. Influence Mechanism Tests

4.4.1. Alleviating Financing Constraints

Following Jiang [50], this study conducted a mediating effect test. Based on the theoretical analysis, the HTEIP can alleviate enterprise financing constraints, thereby encouraging enterprise green innovation. Drawing from the previous literature [51], this study used the SA index as a proxy for financing constraints. Additionally, this study used the end-of-period loan funds (LOAN) to measure a firm’s financing situation. LOAN refers to the funds obtained by an enterprise from financial institutions or other channels through borrowing at the end of the period. These funds are typically used to supplement firms’ liquidity and can effectively alleviate fund shortages.

We add the financing constraint indicators as mediating variables to the estimation equation, and the results are listed in

Table 6. Results of mechanism analysis.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
	SA	LOAN	Gov1	Gov2	RDPerson	Bachelor	RDexp	RDexp_r
Hightech	−0.0038 ***	0.4325 ***	0.0972 **	0.1155 ***	0.0487 ***	0.0129 **	0.0902 ***	0.0717 ***
	(−4.1838)	(4.2264)	(2.4833)	(3.0711)	(5.2728)	(2.3600)	(7.3128)	(6.8111)
Constant	−2.6141 ***	−13.3640 ***	11.6316 ***	12.6952 ***	4.1099 ***	−0.1080	13.3437 ***	2.1773 ***
	(−1.3 × 102)	(−9.2686)	(20.6178)	(23.1127)	(17.2972)	(−1.1552)	(65.2102)	(12.6244)
Controls	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Firm	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Observations	33,561	33,531	21,372	22,214	22,328	28,300	28,505	27,383
R-squared	0.8833	0.7095	0.6730	0.6573	0.8978	0.9292	0.9033	0.8983

Note: t-values in parentheses. ** p < 0.05, and *** p < 0.01.

. Column (1) indicates that the identification policy significantly alleviates firms’ financing constraints. Column (2) suggests that this policy significantly increases firms’ borrowed funds, thereby reducing financial pressure on firms. The reduction in financing constraints directly promotes enterprise eco-innovation [43]. Therefore, it is confirmed that the identification policy encourages firms’ green innovation through its financing constraint alleviation effects, validating H2a.

4.4.2. Government Subsidy Effect

Enterprises that obtain high-tech certification will obtain government subsidies that can facilitate their green innovation activities. We first used the logarithm of government subsidies received by enterprises in the current year (Gov1) and the previous year (Gov2) to conduct empirical analysis. In columns (3) and (4) of Table 6, we can see that enterprises with certification can obtain more government subsidies. Shao and Chen [26] and Xia et al. [35] noted that subsidies promote green innovation. Government R&D subsidies are used to purchase environmental protection equipment, develop green projects, and promote green technology upgrades, thus facilitating firms’ green innovation [25]. Additionally, subsidies can encourage energy-intensive enterprises to engage in eco-innovation [52]. Overall, the high-tech enterprise identification policy through government subsidies provides strong support for enterprises in obtaining more funds in the external capital market. This alleviation of financial constraints is beneficial for enterprises’ green innovation, thus validating H2b.

4.4.3. Talent Agglomeration Effect

Talent is an important foundation for innovation activities. Drawing on Liu et al. [36], we measured the level of S&T talent agglomeration using two indicators: the logarithm of the ratio of R&D personnel to total employment (RDPerson) and the logarithm of the proportion of employees with an undergraduate degree or higher to total employment (Bachelor). The results in columns (5) and (6) of Table 6 show that the intervention of the policy has a significant effect on expanding their R&D teams and enhancing a highly educated workforce. Highly qualified personnel possess extensive knowledge and innovation capabilities and could reasonably assess the risks and feasibility of innovation projects. This helps optimize firms’ green R&D strategies and improve project success rates [53], which can enhance the quality and efficiency of eco-innovation. In summary, the identification policy supports firms’ green innovation by the talent agglomeration effect, which supports H2c.

4.4.4. R&D Investment Effect

The high-tech enterprise identification system increases R&D investment by integrating incentives and self-monitoring mechanisms, thereby promoting enterprise green innovation. The policy requires enterprises to reach a certain proportion of R&D investment over the last three years and stipulates that domestic R&D expenditure must account for at least 60%. This motivates companies to increase their R&D investments continuously. Following previous studies [36], we use total R&D expenditure (RDexp) and R&D expenditures as a proportion of operating revenue (RDexp_r) to reflect firms’ R&D investment intensity. The results are given in the last two columns of Table 6. The coefficients of Hightech for both variables are positive at the 1% significance level, which indicates the identification policy significantly encourages enterprises’ R&D investment.

Increased R&D investment helps enterprises accumulate the necessary technology and knowledge for green innovation, reduces innovation risks and costs, and promotes the successful implementation of green innovation projects [36]. This creates a virtuous cycle of sustainable development and lays a solid foundation for green innovation and long-term competitiveness. In summary, the identification policy boosts green innovation through the mechanism of increased R&D investment; thus, H2d is supported.

4.5. Heterogeneity Analysis

4.5.1. Heterogeneity of Enterprise Ownership

SOEs usually receive more financial subsidies and bank credit support than non-SOEs, and they face fewer financing constraints [54]. Therefore, this study re-performed an empirical analysis by categorizing firms into SOEs and non-SOEs, according to their nature of equity. Columns (1) and (2) in

Table 7. Heterogeneity of enterprise ownership and enterprise scale.

	(1)	(2)	(3)	(4)
	SOEs	Non-SOE	Small-Scale	Large-Scale
Hightech	0.0214	0.0147 **	0.0222 ***	0.0206
	(1.5055)	(2.0240)	(3.1552)	(1.5265)
Constant	−0.3892 **	−0.4181 ***	−0.1971 *	−0.4396 **
	(−2.4560)	(−3.8688)	(−1.8928)	(−2.4966)
Controls	Yes	Yes	Yes	Yes
Firm	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes
Observations	12,254	20,163	15,931	17,203
R-squared	0.6478	0.5998	0.5461	0.6817
Intergroup difference test	−0.059 ***		−0.083 ***

Note: t-values in parentheses. * p < 0.1, ** p < 0.05, and *** p < 0.01.

show that the policy has a significantly positive effect on the eco-innovation in non-SOEs, but no significant impact on SOEs. These results were validated based on intergroup difference tests. The likely reason for this is that SOEs have a substantial advantage in obtaining credit from financial institutions and receiving government support. Thus, the policy’s effect through the financing constraint alleviation mechanism is less pronounced for SOEs. In addition, due to diverse operational objectives and differences in executive compensation levels, SOEs are less likely to invest in R&D and innovation than non-SOEs.

4.5.2. Heterogeneity of Enterprise Scale

Enterprises of different sizes face distinct financing constraints. Due to their strong capital strength and credibility, large-scale enterprises typically find financing easier and have more diversified financing channels to adapt to volatility in the external financing environment. On the other hand, small firms have more limited financing channels and usually rely on traditional financial institutions, such as banks, which results in higher financing costs. The HTEIP provides more opportunities and support for small-scale enterprises, helping them overcome financing difficulties more easily. For large-scale enterprises that may already possess high levels of innovation and financing capabilities, the policy’s stimulating effect is relatively weaker. To investigate whether the effect of the identification policy differs by firm size, this study classified firms according to the median of total assets, with companies above the median classified as large enterprises and the rest as small enterprises. From Column (3) of Table 7, this policy significantly promotes eco-innovation in small-scale companies, while Column (4) indicates that there is no significant promotion effect on eco-innovation in large-scale enterprises. Small enterprises tend to have more autonomy in their operational strategies, lower coordination costs, better alignment with market needs, and a willingness to engage in green innovation.

4.5.3. Heterogeneity of Institutional Quality

Regions with high institutional quality typically have more intense market competition, requiring enterprises to continuously improve their technological and innovation capabilities to gain a competitive advantage. According to Arrow [55], firms in highly competitive environments exhibit stronger innovation motivation. Firms must continuously innovate to survive and maintain a significant share in a competitive market. Moreover, regions with high institutional quality have abundant financial resources, helping enterprises obtain loans more easily. Referring to Wang et al. [56], we divided provinces into low- and high-quality institutional regions on the basis of the median level of marketization. From columns (1) and (2) of

Table 8. Heterogeneity of institutional quality, enterprise factor intensity, and geographical location.

	(1)	(2)	(3)	(4)	(5)	(6)
	High-Quality Institutional	Low-Quality Institutional	Capital Intensive	Labor Intensive	Eastern Region	Western Region
Hightech	0.0192 *	0.0135	0.0282 ***	0.0083	0.0189 **	0.0181
	(1.8984)	(1.3431)	(2.8286)	(0.7235)	(2.4252)	(1.3033)
Constant	0.0067	−0.2918 **	−0.2849 **	−0.4603 ***	−0.3372 ***	−0.3036 *
	(0.0423)	(−2.2441)	(−2.0939)	(−3.2304)	(−3.2807)	(−1.7336)
Controls	Yes	Yes	Yes	Yes	Yes	Yes
Firm	Yes	Yes	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes	Yes	Yes
Observations	16,186	16,644	17,028	14,816	24,211	9350
R-squared	0.7085	0.6226	0.6664	0.6411	0.6405	0.5876
Intergroup difference test	0.022 ***		0.023 ***		0.022 ***

Note: t-values in parentheses. * p < 0.1, ** p < 0.05, and *** p < 0.01.

, we can see that the coefficient of the Hightech is significantly positive only in high-quality institutional regions, while it is not significant in low-quality institutional regions.

This difference occurs because enterprises in high-quality institutional regions are in a relatively competitive environment, and they must continuously innovate to survive and maintain a significant share in a competitive market. Second, regions with high institutional quality usually have a higher demand for green innovation, with consumers placing greater emphasis on innovation in green products and services. Therefore, the incentive effect of this identification policy is more pronounced for firms in regions with high institutional quality.

4.5.4. Heterogeneity of Enterprise Factor Intensity

Capital-intensive enterprises typically rely on advanced production technologies and equipment. The high-tech enterprise identification policy can incentivize firms to increase their inputs in technological R&D, meeting the urgent need for technological upgrades in green production for capital-intensive enterprises. Consequently, capital-intensive firms may be more willing to engage in eco-innovation under the intervention of the policy, which provides more distinct financial support and guidance effects. To test the heterogeneity effect of this policy according to firms’ factor intensity, this study measured enterprises’ factor intensity using the logarithm of the proportion of net fixed assets to the number of employees. Capital-intensive firms refer to those with a factor intensity greater than the median, while the others were considered labor-intensive firms. Subsequently, a subsample regression was conducted separately to discuss the heterogeneity effects.

Columns (3) and (4) of Table 8 show that the coefficient of Hightech for capital-intensive enterprises is statistically significant, while it is not significant for labor-intensive enterprises. It implies that the identification policy significantly enhances green innovation in capital-intensive enterprises. Capital-intensive enterprises typically require substantial funding for green technology R&D and equipment upgrades. This identification policy can alleviate both internal and external financing constraints through external signaling effects and increased internal cash flow. Capital-intensive enterprises face an urgent demand for funding. Therefore, the financial incentive effects of the identification policy are more pronounced, helping promote green innovation within capital-intensive firms.

4.5.5. Regional Heterogeneity

Significant differences exist in the scale and quality of the highly skilled workforce between China’s coastal and inland regions. Labor resources in the eastern regions are more abundant, providing a basis for the talent agglomeration effect of the identification policy. To examine geographical heterogeneity, we classify enterprises into eastern and western firms based on their geographical regions. The final two columns of Table 8 reveal regional heterogeneity in policy effectiveness. The Hightech coefficient is significantly positive for eastern enterprises but statistically insignificant for western enterprises.

This is because the eastern coastal regions of China not only have a higher level of economic development but also possess abundant higher education resources, which can attract more high-quality talent and, thus, maintain a higher overall quality of the labor force [11]. In the eastern region, it is easier for enterprises to hire specialized professionals, resulting in an S&T talent agglomeration effect, which is more helpful to eco-innovation activities. Conversely, China’s western regions face constraints, including limited higher education resources, restricted talent pools, and insufficient S&T personnel for green innovation requirements. These resource constraints diminish the identification policy’s effectiveness in promoting green innovation.

5. Further Analysis

5.1. Firm Life Circle Analysis

Previous analyses of the high-tech enterprise identification policy’s green innovation effects have primarily focused on cross-sectional heterogeneity. Next, this study further analyzed the heterogeneity from the time dimension by dividing firms into different stages according to life cycle theory. Firms at different life cycle stages exhibit heterogeneity in size, profitability, strategic objectives, investment approaches, financing constraints, and innovation propensity [57,58]. Consequently, the high-tech identification policy’s incentive effects may vary across life cycle stages. Following Dickinson’s cash flow model [59], we classify firms’ life cycle stages using net cash flow patterns. To streamline the analysis, we consolidate the introduction and growth stages into a unified “growth stage” category. Similarly, we merge the shake-out and decline stages into a “decline stage” category. This approach provides clearer delineation of firms’ developmental phases and associated financial characteristics.

Table 9. Heterogeneity analysis based on the life cycle.

	(1)	(2)	(3)
	Growth	Maturity	Decline
Hightech	0.0249 **	−0.0066	0.0334
	(2.1834)	(−0.4940)	(1.5350)
Constant	−0.1887	−0.5024 ***	0.0252
	(−1.2997)	(−2.7499)	(0.1168)
Controls	Yes	Yes	Yes
Firm	Yes	Yes	Yes
Year	Yes	Yes	Yes
Observations	14,399	11,289	4939
R-squared	0.6344	0.6975	0.7588

Note: t-values in parentheses. ** p < 0.05, and *** p < 0.01.

presents the estimation results, revealing significant heterogeneity in policy effectiveness across life cycle stages. The identification policy demonstrates stronger incentive effects on green innovation for growth-stage firms while showing no significant impact on mature and declining enterprises.

Generally, growth-stage firms often focus on innovation activities and reinvest in and expand their businesses to enhance market competitiveness [60,61]. Their willingness to participate in green innovation is also stronger. Through green innovation, firms can gain core competitiveness, which helps increase their market share and meet the strategic development needs of growth-stage enterprises. However, in the face of the high risk and uncertainty associated with green innovation, growth-stage firms may face resource constraints, such as financial limitations and shortages of specialized talent, which can affect their green investment [61]. Through tax incentives and government subsidies, the high-tech enterprise identification policy effectively reduces operating costs and provides enterprises with more funds for R&D investment in green innovation. Moreover, it promotes talent agglomeration by attracting more professional talent to growth-stage firms. This expands the team size and establishes a solid human capital foundation for the R&D of green technologies.

Mature firms face a series of challenges and constraints that make green innovation relatively less favorable at this stage. Consequently, the high-tech enterprise identification policy does not have a significant impact on promoting eco-innovation in mature firms. First, maturity-stage firms typically have strong financial stability and high profitability, and have established stable operational models and market positions. This stability can lead to resistance to changes and innovations [61]. Managements may prefer conservative business strategies and be averse to the risky nature of green innovation. Second, maturity-stage firms have already established a series of business relationships and interest systems, including reliance on traditional supply chains. Promoting green innovation may involve conflict with existing stakeholders, creating additional difficulties for enterprises pursuing green innovation initiatives. Third, maturity-stage firms typically have a relatively stable market share and face less competitive pressure. Green innovation often has a long investment return cycle, and mature firms may prioritize short-term profit stability over strategies with long-term investment returns. Therefore, the high-tech enterprise identification policy provides less incentive for green innovation in maturity-stage firms.

Similarly, enterprises in the decline stage often experience a reduction in market share and profitability, which presents a series of challenges for green innovation [62]. Decline-stage firms typically face financial pressures, such as cash shortages and debt problems. In this case, these firms may prioritize survival and coping with current economic pressures over investing in green innovation. Additionally, firms in the decline stage often implement austerity measures to sustain their operations, which can decrease employee motivation for innovation. Simultaneously, these firms also face talent loss, where employees may lack the enthusiasm to participate in green innovation. As a result, a favorable internal foundation for green innovation is lacking within decline-stage firms, and the incentive effects of policy are weak.

5.2. Green Innovation Efficiency

This study further explored whether the identification policy affects the enterprise green innovation efficiency. Previous studies have adopted data envelopment analysis (DEA) and stochastic frontier analysis (SFA) to measure green innovation efficiency [47,63]. However, these measurement methods are often used at the macro-level. At the firm level, this paper constructed an index of enterprises’ green innovation efficiency using the ratio of green innovation output to green innovation input.

Eco-innovation output is measured as the logarithm of total green patent applications plus one, while eco-innovation input is captured by the logarithm of annual R&D expenditure plus one.

Table 10. Results of green innovation efficiency.

	(1)	(2)
Hightech	0.0093 ***	0.0049 ***
	(17.0465)	(7.2422)
Constant	0.0200 ***	−0.0275 ***
	(49.5686)	(−2.7353)
Controls	NO	Yes
Firm	NO	Yes
Year	NO	Yes
Observations	29,367	28,470
R-squared	0.0098	0.7047

Note: t-values in parentheses. *** p < 0.01.

presents DID estimation results examining the impact of HTEIP on green innovation efficiency.

Column (1) reports baseline results excluding fixed effects, incorporating only the core explanatory variable Hightech. The Hightech coefficient (0.0093) is statistically significant at the 1% level. Column (2) demonstrates that this coefficient retains significance after including all control variables. These findings indicate that the HTEIP significantly enhances firms’ green innovation efficiency and promotes comprehensive eco-innovation development.

6. Conclusions and Policy Implications

6.1. Conclusions, Limitations, and Further Research

The high-tech enterprise identification policy is a strategic initiative in China aimed at breaking through critical technological barriers and leading industry advancements. It is worthy of in-depth study to investigate whether it promotes enterprises’ green innovation within the context of sustainable development. Using data from China’s A-share listed companies from 2008 to 2022, this study employed a time-varying DID method to analyze the impact of the identification policy on corporate green innovation and its mechanism. The study reveals that the identification policy significantly promotes green innovation. The conclusion remained valid after robustness tests. The key channels driving this effect include alleviating financing constraints, obtaining more government subsidies, promoting the agglomeration of S&T talents, and increasing enterprises’ R&D investments. The heterogeneity analysis indicated that the promotional effect is stronger in non-SOEs, smaller firms, capital-intensive enterprises, firms from regions with high institutional quality, and China’s eastern regions. Additionally, growth-stage firms demonstrated a stronger willingness for green innovation than mature and declining firms, making the incentivizing effect of the identification policy more pronounced for growth-stage firms. Furthermore, this policy can significantly enhance green innovation efficiency.

While this study provides valuable insights, we acknowledge several limitations due to certain data and methodological constraints. First, it relies on data from Chinese listed firms, which may not fully capture the dynamics of small and medium-sized enterprises (SMEs). Moreover, this study focuses on China’s institutional context, limiting its applicability to other economies. Second, green innovation is measured using patent data, which may underestimate non-patented or informal eco-innovations. Third, while the DID framework addresses endogeneity to some extent, potential biases from unobserved confounders may remain.

These limitations also point to directions for subsequent research. Future investigation could explore cross-country comparative analyses, examine the long-term dynamic effects of such a type of selective industrial policies. In addition, future research could incorporate qualitative survey data to better capture the innovation strategies and behavioral responses of SMEs. From the perspective of methodology improvement, subsequent research could adopt advanced causal inference methods—such as causal forests or double machine learning—to obtain more robust and unbiased estimates.

6.2. Policy Implications

The following policy implications are derived based on the findings of this study. First, it is essential to refine the high-tech enterprise identification policy to ensure the in-depth implementation of enterprise green innovation. The policy should include detailed provisions and standards to promote green innovation and establish a sound green innovation policy framework that includes tax incentives, subsidies, and reward measures to encourage firms to increase their R&D inputs and adoption of green technologies. Second, financing constraints have always been a common challenge for high-tech enterprises implementing green innovation on a global scale, particularly in developing countries. To mitigate these financing pressures, governments can establish dedicated green innovation funds or guide international collaboration to increase their investments in eco-innovation. Third, high-tech talents are essential for green technological innovation. Governments should attract and retain more global talent by relaxing their visa policies, providing generous work benefits, and building international R&D platforms.

Fourth, policies should be tailored to the specific needs of enterprises and regions to better leverage the benefits of high-tech enterprise identification policies. For example, non-SOEs should be ensured a fair and competitive environment and be provided with the necessary policy support and incentive mechanisms. For small-scale enterprises, simplifying approval processes, offering technical support, and providing exchange platforms can lower barriers to green innovation. For firms in high institutional quality regions and capital-intensive enterprises, they should receive more green investment incentives and financing support to drive green transformation and strengthen risk management. The eastern regions should serve as exemplary leaders, enhancing regional cooperation, promoting green lifestyles, and increasing public participation. Finally, according to the life cycle theory of enterprises, a differentiation strategy should be employed to further refine the standards of the identification policy. Thus, to support growth-stage enterprises, governments should adopt more targeted measures, such as moderately relaxing identification standards, improving complementary policies, and increasing financial support. Also, it is particularly important to be vigilant against and prevent potential speculative behaviors and R&D manipulation by enterprises.