Configurations of Technology Commercialization: Evidence from Chinese Spin-off Enterprises

Despite numerous studies on understanding effectiveness of technology commercialization, only limited attention is given to the complex relationship inter and inner its conditions, which formed from the policy system between different levels of government networks and the unbalanced regional technology resources in China. Using data from spin-off enterprises, this paper develops an integrated configuration framework for and provide an empirical test of five predictors, namely policy supply, policy coordination, technology source, firm strength, and R&D investment, which capture three contextual features of institutional, organizational, and technological complexity within Chinese technology commercialization context. Confirmatory factor analysis (CFA) and Fuzzy-set qualitative comparative analysis (fs/QCA) of data identified three distinct effective TC paths: The first, dominator, pertained to the effective regional policy supply and powerful enterprise strength, even if the regional technology source is deficient. The second, devotee, characterized by effective policy coordination, intensive R&D investment, and abundant technology source or powerful enterprise. The last, investor, featured by poor policy coordination and poor regional technical resources, while favorable regional policy supply and R&D investment.

But in spite of the large number of studies on technology commercialization (TC) in recent years, there are two empirical challenges that need to be resolved towards the understanding of TC effectiveness.First, studies have focused a lot on single theoretical perspectives (Cunningham & Paul, 2018;Good et al., 2019) with downplaying the joint effects of multiple perspectives and the synergistic, interactive as well as the interrelationships conditions, named, the integrated perspective, which perceived the performance of TC a synergy of the characteristics of various conditions (Etzkowitz & Leydesdorff, 2000;Leydesdorff, 2011;Bozeman et al., 2015;Carayannis & Campbell, 2012;Carayannis et al., 2018).More specifically, the literature has discussed and tested a number of different factors or perspectives that drive effective TC.Some of these are related to the resource, such as technology infrastructure, like university (Audretsch et al., 2012;Etzkowitz & Leydesdorff, 2000;Guerrero & Urbano, 2012).Others reflect the institutional context in which the effectiveness realizes, such as TC policy supply (Li et al., 2015;Ergas, 1987;Perkmann et al., 2013), TC policy coordination (Du & Wang, 2019;Huggins and Thompson, 2017;Zheng et al., 2015).Yet others discuss the organizational elements that facilitate enterprises' TC, like firm strength and R&D investment (Meyer & Goes, 1988;Damanpour & Evan, 1984;Hitt et al., 1991;Cohen & Levinthal, 1990;Pandza & Holt, 2007).To the extent that all these factors or perspectives matter individually, any empirical analysis that ignores some of them provides inadequate explanation.In this regard, it is necessary to conduct empirical analysis with them collectively as configurations rather than fragmentary predictors.
The second challenge relates to the regional adaptability of existing integrated frameworks.In this point, previous TC study pays much attention to European and US regions though less to emerging economies, such as China (Frondizi et al., 2019;Hou et al., 2019).Whereas, the theoretical applicability of TC arrangements varies from region to region, due to influences like cultural factors, geographic proximity, and regional institutional regime factors (Audretsch, 2014;Clarysse et al., 2005;Cunningham et al., 2018;Jefferson et al., 2017;Kantis et al., 2004).Given the discrepancy of policy systems among government networks at all levels and of regional technological resources as well as its distinct humanistic background in China, there are even more complex condition sets for TC effectiveness in the country (Huang et al., 2013;Lu & Wang, 2012).To make sense of such effectiveness, it calls for an explicit reorganization and empirical analysis within Chinese policy context with an integrated framework.
In this paper, we address these challenges in order to develop an empirical understanding of the configurational effectiveness of TC in China.First, we view

Integrated TC Framework
The technology commercialization (TC) is one of the key issues in the field of innovation management (Christensen, 2013;Cai, 2015).Similar concepts refer to technology transfer or technological achievement transformation.In fact, technology management scholars have been exploring technology transfer in depth for more than 30 years (Zhao & Reisman, 1992;Carayannis & Campbell, 2009, 2012;Kirchberger & Pohl, 2016).Zhao and Reisman (1992) categorized technology transfer research into three perspectives: process perspective, strategy perspective, and alliance perspective.In the newest review articles, Kirchberger and Pohl (2016) summarized TC research into organizational perspective, resource perspective, and institutional perspective.Although these theoretical perspectives were proposed in different names, common categorical basis are shared: the perception of institution, resources, organizations, as well as managers to the TC effectiveness.
There are two main streams of grounding for TC research: one is the single perspective, either taking universities as grounding points to study university-industry technology transfer (Chen et. al, 2016;O'Shea et al., 2008;Perkmann et al., 2013); or taking intermediaries as grounding points to study importance of intermediary in TC process (Howells, 2006;Bigliardi & Dormio, 2017;Battistella et al., 2016); or taking spin-off firms as grounding points to verify the impact of firm characteristics on TC effectiveness (Feldman & Audretsch, 1999;Glaeser et al., 2002;Huang et al., 2013;Guo et al., 2021).Second, integrated perspectives, e.g., based on the triple helix framework of industry-academia-government (Etzkowitz & Leydesdorff, 2000;Gunasekara, 2006), or the quadruple helix framework (added social innovation and public value dimensions) (Leydesdorff, 2011;RIS, 2014) and the quintuple helix framework (added environmental conditions once more) (Carayannis & Campbell, 2012;Bozeman et al., 2015), as well as the entrepreneurship perspective, which, argues that TC occurs in an ecosystem containing university, government, industry, society, and the environment that including a wide range of interconnected players (Carayannis et al., 2018;Good et al., 2019).
The integrated TC framework provides a systematic and comprehensive perspective to fully understand TC performance with capturing features of three conditions, namely institutional condition, technological resources, and organizational behavior.However, the interaction and configuration effects between institution supply and institution coordination factors under particular resource and organizational background and the interrelationship among on TC performance have not been empirically tested enough.Especially within the contextual of Chinese spin-off enterprises.

Conditions for TC Performance in Spin-off Enterprises
Rooted on integrated framework of TC, this paper seeks to identify the causal conjunction both within and across the five conditions by conducting Fuzzy-set qualitative comparative analysis (fsQCA) (Ragin, 1987) of the TC performance of 186 spin-off firms.Since the above conditions cover three dimensions as institutional, organizational, and resources, and each contains a variety of specific influencing conditions, it is unrealistic and difficult to take all conditions into consideration.Therefore, with identifying the crucial conditions according to deductive and inductive approach, five conditions, namely, supply and coordination of policy (institution level), technology source (resource level), strength of science and technology enterprises, and R&D investment (organization and manager level), were adopted 1 3 Journal of the Knowledge Economy (2023) 14:3453-3488 to investigate the configuration effect of their interactions on high-performance TC.Following is the influence mechanism analysis of five conditions on high-effective TC.

Policy Supply and TC Performance
The seminal work of Baumol (1990) emphasizes the importance of institutional factor in promoting and shaping entrepreneurship as well as innovation outcomes.One of the critical role that institutions play is to encourage and facilitate TC by setting norms that guide entrepreneurial activities (Li et al., 2015).This can be interpreted, on one hand, task-oriented policies guide direction for technological development and technology integration for industries and technology-producing organizations (e.g., universities), while diffusion-oriented policies sponsor broader technological development at more advanced stages of product life cycle by building common technology infrastructures (Ergas, 1987;Conle et al., 2021).On the other hand, policy makers promote TC effectiveness by enacting TC policies which allocated funds to intellectual properties, such as incentives and fiscal and tax policies (Geoghegan & Pontikakis, 2008;Perkmann et al., 2013;Wonglimpiyarat, 2016).
Geographic proximity was considered as a factor influencing institutional impact.For instance, technology-based firms which located in same country or region tend to be more similar as sharing common institutional regime (Jefferson et al., 2017).Moreover, regional institutional regime in which entrepreneurial universities operated in facilitate entrepreneurship, innovation, and value creation in the regions (Cunningham et al., 2018).Nevertheless, as observed by researches, there are more policy discrepancy for TC across regions in China (e.g.: Huang et al., 2013;Lu & Wang, 2012), which made policy supply one of the conditions in the integrated framework.

Policy Coordination and Tc Performance
Well-formulated TC policies should be established in bundles of main policies and flexible functional policies (Du & Wang, 2019).The coherence of policies oriented to diverse target organizations, government levels, and geographic regions is fundamental for realizing research cooperation between firms and external technological resources, which consequently stimulate innovation activities in the region.The interaction of various different entities, i.e., dynamics network, ultimately triggers intra-and inter-regional knowledge flows that lead to innovation and economic growth (Huggins & Thompson, 2017).
Changing environment was found an important factor influencing the process of policy implementation.TC policies that matching and adapting quickly to the dynamic environment is vital to effectiveness.Random multi-path information feedback mechanisms (Zheng et al., 2015) was mentioned avoiding poor communication and information asymmetry which frequently occurs between TC operators like policy-maker, implementer, and target organizations, and may lead to poor communication or misinterpretation.Meanwhile, policy implementation rooted in particular socio-cultural background might enhance effectiveness of TC due to better guarantee the legitimate rights and meet the demands of diverse operators (Shang & Yang, 2013).

Enterprise Strength and TC Performance
There is a consensus that organizational size affects the innovation process in organizations (Kimberly & Evanisko, 1981;Meyer & Goes, 1988).Economies of scale associated with organizational size increase the feasibility of technology adoption (Nord & Tucker, 1987).Big organizations normally have more opportunities to access complex and diverse resources as well as capabilities.The larger number of skilled personnel (Damanpour & Evan, 1984) and technical know-how ensures firms to adopt additional innovations be involved in larger volume of business thereafter affording more business risks (Hitt et al., 1991).Also, bigger-scale organizations are more capable at tolerating more innovation loses, or be involved in larger volume of business thereafter affording more business risks (Hitt et al., 1991).However, it is often argued that fundamental organizational contextual characteristics such as firm age and enterprise size often lead to complex administrative structures and internal inertias (Cohen, 1995;Camiso et al., 2004), which tend to hinder the coordination and communication of external learning therefore reduce adaptability and cause TC ineffectiveness.

R&D Investment and TC Performance
Continuously allocating capital on internal R&D activities or purchasing external technologies are commonly used approaches firms applied to realize innovation.Both approaches involve R&D investment, also named as intensity of R&D or R&D inputs.Although measurement criterion varies, the significance of R&D investment on innovation performance has been empirically confirmed in many studies.Generating knowledge and inventions and improving firms' ability to absorb and utilize external knowledge were major mediators in firm's absorptive capacity (Cohen & Levinthal, 1990).Managers frequently develop firm's absorptive capacity by internal R&D investment and external resources linkage to build knowledge bases (Cockburn et al., 2000), which in turn determines the capability in which their enterprises draw external knowledge from competitors, government, university laboratories, and other outside industrial sources, so as to facilitate innovation (Pandza & Holt, 2007).

Technology Sources and TC Performance
The importance of universities in regional development and innovation-driven growth was pervasively recognized by economists and policy makers (Audretsch et al., 2012).In knowledge-based economies, universities and government research centers are public of knowledge-producers, providing highly skilled human resources and critical knowledge resources for innovation (Etzkowitz & Leydesdorff, 2000).Simultaneously, it acts as disseminators, directing knowledge and technology flows, connecting talents, firms, and collaborations together (Guerrero & Urbano, 2012 & Etzkowitz, 1996), and supporting entrepreneurs in knowledge transfer, incubation, mentoring, and consulting.Consequently, stimulate generations of increasing number of entrepreneurial as well as innovative outcomes (Bercovitz & Feldman, 2006;Salter & Martin, 2001).Numerous studies indicated that knowledge spillovers mostly occur within geographically scoped regions, as proximity ties are more likely to foster innovative impulses than distant ties.Meanwhile, focal firms closer to universities significantly exhibit higher level of innovative activities (Audretsch & Lehmann, 2005;Feldman & Audretsch, 1999;Glaeser et al., 2002;Jaffe et al., 1993).
However, a number of scholars disagree this, arguing that academic knowledge of universities was not clearly used geographic locally (Brostrom, 2010;Zucker & Darby, 2001), backing by the evidence that highly innovative companies, which at the forefront of academic knowledge in their industries, usually need to collaborate cutting-edge technologies global-widely (Monjon & Waelbroeck, 2003), so that often success in establishing international research networks to gain access to outstanding research contexts (Laursen et al., 2011).
The "configuration approach" holds that in process of complex organization governance, multiple conditions are interdependent with each other and tend to combine into differentiated permutations in achieving common organizational outcomes (Du & Jia, 2017;Fiss, 2011).From this perspective, the paper empirically explored the up mentioned conditions and the configuration paths of effective TC in the practical scenario of spin-off enterprises.A synthesis model with the relationships developed in the literature review was proposed in Fig. 1.

Qualitative Comparative Analysis (QCA) Method
An empirical test using fs/QCA method was proposed for this study.Fs/QCA approach was put forward by Ragin (1987), initially been applied to political science and sociology research, more recently, been applied increasingly frequent in management science (Greckhamer, 2016).Given the research problem, firstly, the QCA method helps to explore multiple concurrent relationships between multiple conditions.Secondly, it can identify different causal paths of high-performance TC, in other words, multiple configurations will be formed between diversified conditional groups each associated with high-effective TC.Finally, QCA method inferred the causal relationship between conditions and results through sets relation instead of correlation, which in line with the actual social phenomenon of the research problem.
From the perspective of set theory, QCA method can identify which conditions (configuration) are necessary or sufficient conditions (configuration) of the result.According to data types, QCA can be divided into csQCA (clear-set qualitative comparative analysis, binary data), mvQCA (multi-valued-set qualitative comparative analysis, multi-valued data) and fsQCA (fuzzy-set qualitative comparative analysis, continuous data).Considering that fs/QCA can further process changing degree or partial membership of data (Du & Jia, 2017), this paper chooses fsQCA as the research method.

Sample Selection and Data Sources
The data used to conduct empirical test mainly comes from questionnaire collection.More specifically, data collection was carried out in two ways: Firstly, contact the local Economic and Information Department (Committee), SME Service Center, Science & Technology Park Management Committee, and park developers in the four regions of the research program (Guangdong, Hubei, Shanghai, and Zhejiang) respectively, enquiry about lists of contact information of TC-related organizations located in the park (intermediaries/enterprises).Through screening and sampling, about 1200 out of a totally 30,000-40,000 organizations were randomly selected for contact in each region, according to a 30-select-1 method.Preset questionnaire websites were sent to about 900 organizations (intermediaries/enterprises) which gave feedback.Relevant personnel was asked to fill in the electronic questionnaire.Secondly, a small portion of questionnaires was collected by using "acquaintance relationship" and various research opportunities directly interview managers of TC intermediaries/enterprises and request for filling electronic questionnaires on site or online.The number of the second method questionnaire was about 20.Nine hundred nine questionnaires were distributed through up-mentioned two methods.Contents of the questionnaires mainly included the strength of TC intermediaries/enterprises, the evaluation of regional TC policy supply and policy coordination, and the evaluation of R&D investment and the TC performance.Among the valid questionnaires collected, 186 samples were from spin-off enterprises, which contained 66 valid questionnaires of Guangdong Province, 29 of Hubei Province, 49 of Shanghai city, and 42 of Zhejiang Province.The overall recovery rate was around 20%.Five-point Likert scale was used in questionnaires.

Measurement and Calibration of Outcome Variable and Antecedent Variables
Outcome Variables TC Performance.Firstly, in line with the existing studies, "quantity of innovation," namely, the number of patents (including claimed) and the number of new products, were used to measure innovation output of spin product enterprises (representing the dominant output in the technology development phase) (Alvarado et al., 2017).Since TC pertains to transforming new ideas into practice, the "output value of new product" indicator was assessed simultaneously to evaluate the knowledge commercialization stage performance of firms (representing the dominant output in the commercialization phase) (Li et al., 2019).

Conditional Variables
Policy Supply.Taking stock of existing literature (Ergas, 1987;Link, 2010), the supply characteristics of TC policy were portrayed in four aspects: directionality, specificity, supportive, and systematicness.Since available data were of third-level indicators instead of second-level, all indicator dates were extracted by confirmatory factor analysis (CFA) to secondary level.Sufficient resource provision, well-established strategic technology R&D infrastructure, well-established common technology R&D infrastructure, and technology diffusion system were extracted as one secondary indicator.Policy focuses on specific industries (or specific product life cycle stages) or specific technologies, and emphasizes supporting emerging industries was extracted as another secondary indicator.Clear policy objective, sufficient financial assistance like funds, subsidies, and taxes, as well as adequate monetary sponsor was extracted as the third secondary indicators.(Appendix Tables 8, 9, 10, 11).
Policy Coordination.Refer to Li and He (2009), Du andWang (2019), andZheng et al. (2015), the coordination characteristics of Chinese TC policy were described in four dimensions: the flexibility of policy adaptiveness, the complementarity of policy implementation, the degree of feedback to participation, and the degree of scientific cultural background.Similarly, as there is only third-level indicator data available, four secondary indicators were extracted by confirmatory factor analysis (CFA): First, degree of policy propaganda, degree of public participation in policy formulation, degree of public participation in policy implementation, degree of expressing public willingness in policy formulation and implementation processes, and degree of adoption, smooth, and responsiveness of various feedback channels.Second, the coordination degree of policy contradictions between different government levels, the coordination degree of policy overlap at different government levels, the degree of timely and proper handling of policy contradictions, the degree of policy learning within organizations, and the degree of public understanding of newly promulgated TC policies.Third, the evaluation of availability and cultural adaptability of TC policies.Fourth, the degree to which policies could be adjusted according to changes in supply and demand in marketplace.(Appendix Tables 12,  13, 14, and 15).R&D Investment.Assessed in two indicators: the degree of significance be attached to R&D by enterprise leaders, and the degree of incremental R&D investment in the past 3 years.These indicators reflect the role of managers in organizational behavior.
Enterprise Strength.Four indicators (enterprise scale, establishment years, number of technicians, and employee education background (Jefferson et al., 2017)) are used to describe the strength characteristic of TC enterprises, reflecting the role of organizational resources.
Technology Sources.The latest announced numbers of university located in four regions were used to describe reflect this indicator.The numbers of local R&D institution (Chen et al., 2016) were also considered.
The mean values of the above indicators were used for subsequent analysis.Measurement of each variable was attached in Appendix 1 and Tables 6, 7.

Calibration of Variables
Calibration refers to converting a raw score into one that reflects degree of membership in a set, rescaling the original measure into scores ranging from 0.0 to 1.0 (Ragin, 2008), the two ends signifying the qualitative thresholds of full membership and full non-membership.Direct calibration approach was used in this research, which converts raw score on three qualitative anchor points through logical functions.Three anchor points are 1 (full membership), 0.5 (intersection), and 0 (full non-membership) (Ragin, 2008).Consistent with prior studies, calibration points of policy supply, policy coordination, and R&D investment, as well as high-performance TC were set at the figure of "mean value surplus standard deviation," "mean value," and "mean value minus standard deviation." Calibration points of enterprise strength variable were set at the figure of "minimum value," "mean value," and "maximum value" respectively.For the technology sources variable, following mainstream QCA study (Greckhamer, 2016), using quantile value to determine three anchor points, i.e., 10-percentile value, 50-percentile value, and 90-percentile value respectively (Table 1).

Analysis Process
From the perspective of set theory, the fsQCA method is aiming at identifying sufficient or necessary subset relationship between antecedent variables and their configurations and outcome (Ragin, 2008).A variable (or configuration) is a necessary condition for an outcome if it always present when the outcome occurs.Therefore, the result is a subset of the condition (or configuration); if the result always occurs when a variable (or configurations) exists, then the variable (or configuration) is a sufficient condition (or sufficient configuration) for the result.In this case, the condition (or configuration) is a subset of the result.Schneider and Wagemann (2012) argue that a believable practice in QCA analysis is to analyze both necessity and sufficiency, and necessity analysis precedes sufficiency analysis.Since sufficient configuration is perceived as core part of QCA analysis and attracts more interesting of researchers, QCA analyses mainly focus on the necessity of single condition and the sufficiency of condition configurations.In this view, the paper firstly analyzes the necessity of five antecedent variables and their "non-set" status for high-performance TC.As widely accepted in previous studies, consistency threshold of 0.9 was used to judge necessity in this paper.More specifically, when consistency of a specific condition is greater than or equal to 0.9, the condition is considered necessary for the result; when consistency of a specific condition is less than 0.9, the condition is considered unnecessary for the result.Secondly, carry out sufficiency analysis, which used the truth table algorithm to determine the sufficiency of relevant configuration and the outcome, also measure with consistency.Ragin (2006) argues that when the consistency between a particular configuration and the result is greater than or equal to 0.75, then configuration could be considered sufficient for the outcome.According to the gap of consistency score in the truth table (Schneider et al., 2010), consistency threshold was determined at 0.85.In addition to the consistency threshold, the number of cases covered by a specific configuration is also the screening criterion for a specific configuration to enter the Boolean minimization process.We set a frequency threshold that specifies the minimum amount of cases that will be considered in the analysis.Usually, the determination of frequency threshold depends on the sample size.Generally, the larger the sample size, the greater the frequency threshold.For small and medium-sized samples (about 10 ~ 100 cases), the frequency threshold usually not less than 1.While for large quantity samples, this value can be appropriately increased (Schneider & Wagemann, 2012).Considering sample size of this paper, the threshold of frequency was set to 1, at which more than 75% of samples were included in the analysis.Finally, the robustness method of set theory was applied to assess the sufficiency of configurations (Zhang & Du, 2019).
The minimization program of QCA 3.0 produces three solutions based on different simplified assumptions: complex solution (without logical remainder), parsimonious solution (with all logical remainder, regardless of whether it conforms to theoretical and practical knowledge) and intermediate solution (with only logical remainder conforming to theoretical and practical knowledge).As Ragin (2008) argues, in general, intermediate solutions are superior to complex and parsimonious solutions because of achieving the balance between them in complexity, it is a combination of theoretical and empirical complementarity (Schneider & Wagemann, 2012).In view of previous inconsistent research conclusions about the relationship between the five antecedent variables and high-performance TC, no clear counterfactual analysis has been conducted in analysis.Following mainstream practices (Fiss, 2011;Ragin & Fiss, 2008), the intermediate solution were mainly presented with supplementation of the parsimonious solutions.The conditions presented in both intermediate solution and parsimonious solutions were viewed as core conditions, while the conditions presented only in intermediate solutions were viewed as peripheral conditions.
Evaluation of above solutions were essentially followed by two metrics: consistency and coverage.Consistency reflects the degree to which a specific solution or all solutions are subsets of the results.Coverage reflects the extent to which the result can be interpreted by a specific solution or all solutions.Coverage can be classified into three categories as: raw coverage, unique coverage, and solution coverage.Raw coverage refers to the degree to which the result can be interpreted by a specific configuration, including the part be interpreted jointly with other configurations.Unique coverage refers to the degree to which the result can be interpreted by one specific configuration only, excluding the part be interpreted jointly with other configurations.Solution coverage refers to the extent to which the results are interpreted by all configurations.

Necessity Analysis of Individual Conditions
It is necessary to examine first the necessity of individual conditions one by one before analyzing the condition configuration.Necessity refers to the condition that must exist for an outcome to occur, but its existence does not necessarily lead to the occurrence of the outcome.Table 2 shows the necessity analysis results of the five antecedent conditions for high-performance TC.All consistency values are found lower than 0.9, indicating no condition constitute a necessary condition for the result, implying the necessity to investigate influence of condition configuration on high performance.

Sufficiency Analysis of Configuration Conditions
FsQCA3.0 software was used to analyze the conditional configurations that lead to high-efficiency and non-high-efficiency TC.Different configurations represent different conditional conditions that achieve the same result (high-efficiency or nonhigh-efficiency TC).The configurations found were named according to the theoretical process of configuration (Zhang et al., 2020).
Table 3 shows the configuration analysis results of achieving high-level performance, presented in the form proposed by Ragin (2008), where • indicating the presence of conditions, that is, the value of condition variables is higher, the big circle represents "core condition," and the small circle represents "peripheral condition;" ⊗ represents condition absence; and blank represents irrelevant condition.
Five configurations were observed leading to high-performance TC, with an overall consistency of 0.837.For individual configuration, the consistency values are of 0.89, 0.88, 0.85, 0.88, and 0.88 respectively, which exceed the common acceptance consistency standard of 0.80, signifies that more than 80% of all cases of spin-off enterprises that satisfy these five conditional configurations exhibited higheffectiveness.The overall coverage of the solution is 0.62, which signifies the five conditional groups can explain 62% of high performance of TC enterprises.Both the consistency and coverage of solutions are higher than critical values, indicating the validity of empirical analysis.

Configurations of High-Performance TC
We set the original consistency threshold at 0.85 and the case frequency threshold at 1. Identified each core conditions by comparing the nested relationship between the intermediate solution and the parsimonious solution: the conditions appearing in both intermediate solution and configurations solution are core conditions, and the condition appearing only in the intermediate solution are peripheral conditions (Du & Jia, 2017).Interpretation of TC configurations 1 to 5 in Table 3 are as follows: Nineteen cases were found in this configuration.A study of these cases showed that 10 of them 19 were located in Shanghai City and the other 9 in Zhejiang Province, which consistent with two facts: (i) Shanghai is the national pilot for TC policy supply, ii) Zhejiang is geographically closed to Shanghai and go speedily after Shanghai in TC policy supply.This configuration indicated that given certain regional policy supply condition, spin-off enterprises can achieve highperformance TC as long as it has strong strength.2. Policy-based firm strength model.Partly same to configuration 1, configuration 2 consists efficient policy supply and powerful enterprise strength as core conditions, efficient policy coordination as the peripheral condition.
Reflects institutional and organizational conditions as same as configuration 1. Sound policy conditions here include not only efficient policy supply but also efficient policy coordination, embodied in: (i) extensive policy propaganda by experts and scholars, high degree of public participation, and public willingness expression in TC policy formulation and implementation, also high in degree of adoption and smooth responsiveness of various feedback channels; (ii) efficient coordination of policy contradictions between different government levels (e.g., contradictions in policy goals and specific regulations), efficient coordination of policy overlap at different government levels (e.g., overlapping in policy functions and targets), the degree of timely and proper handling of policy contradictions (e.g., talent technology support, science, and technology financing, fiscal policy), frequently policy learning within organizations, and high degree of public understanding on new policies; (iii) TC policies are accessible and high in cultural adaptability; (iv) timely adjustment of TC policies according to changing supply and demand marketplace.Configuration 2 suggests that given sound regional TC policy condition, TC firms can achieve high-performance TC if they are powerful in firm strength.Nineteen cases were found in this configuration.Eight among which were in Shanghai, six were in Guangdong Province, and two and three in Hubei and Zhejiang respectively, indicated universality of this configuration.
3. Policy-coordination-based R&D model.Configuration 3 consists efficient policy coordination, high intensity R&D investment, and rich regional technological resource as core conditions.This configuration exhibit a triad of institutional, organizational, and technological resources matching.Heavy R&D investment evidenced by two indicators: (i) importance was attached to R&D investment by enterprise leaders; (ii) significant increase in R&D investment fund in the past 3 years.Technology resource condition reflected in abundance of regional technological resources.The configuration indicates that, in regions where have wealth in technological resources, TC enterprises can obtain high-performance TC only if the coordination of existing TC policies is efficient, and the enterprises' R&D investment is sufficient, no matter efficiency of policy supply.Interestingly, all of the 19 cases in this configuration was found in Guangdong Province.Consistent with two facts: (i) Guangdong Province has 154 universities, the highest among these four regions ② , (ii) Guangdong Province has the highest technology market turnover in 2019 (222.308 billion yuan) among four regions.4. Policy-coordination-based R&D plus technology model.Configuration 4 consists efficient policy coordination, heavy R&D investment and powerful enterprise strength as core conditions.It also exhibits the characteristics of institutional, organizational, and technological resources matching.Suggests that under general regional policy supply, high effectiveness TC can be produced only if the coordination of existing policies is effective, and the enterprise is powerful in strength and invest sufficient capital on R&D.This configuration corroborates and extends existing research that larger organizations have more complex and diverse resources and capabilities, such as more skilled personnel and more technical know-how, enabling itself to adopt more innovations (Hitt et al., 1991).Meanwhile, the higher educational background of scientific and technical personnel is also one of the strength factors to promote TC.Compare with configuration 3, cases in configuration 4 have appeared in all the four regions involved in the investigation, including Zhejiang, Hubei, Guangdong, and Shanghai.Among which, six are from Shanghai, nine from Guangdong, and two from Hubei and Zhejiang respectively.As same as configuration 2, configuration 4 is also a universal combination and need to be noticed.5. Policy-supply-based R&D model.Configuration 5 consists policy supply efficiency, intensive R&D investment, inefficient policy coordination, and scare of regional technology resources as core conditions.This configuration exhibits the matching feature of institutional and organizational elements (managerial behavior).The path featured by lack of policy coordination but powerful in policy supply and R&D investment, which corroborates and extends existing research that managers can improve innovation performance by investing in internal R&D and establishing links with external resources to build a knowledge reservation (Cohen & Levinthal, 1990).
The number of case in this configuration is relatively few, totally six.Among them, five cases were found in Zhejiang province and one in Shanghai city, consistent with the fact that Shanghai and Zhejiang are comparatively abundant in policy supply, while inferior in technological resources (number of university) in the four regions.
A comprehensive comparison further reveals the interaction between conditions.Firstly, both configurations 1 and 5 contain policy supply condition while lack of regional technology source.The condition of powerful firm strength in configuration 1 and heavy R&D investment in configuration 5 suggests that R&D investment (managerial behavior) has some substitution for firm strength (organizational conditions).Secondly, both configurations 2 and 4 contain policy coordination as well as firm strength condition, but configuration 4 consists intensive R&D investment condition while without efficient policy supply, indicate that R&D investment (managerial behavior) has a substitution effect on policy supply (institutional condition).Thirdly, both configurations 2 and 5 contain policy supply conditions, configuration 2 consists of powerful firm strength and better regional policy coordination, while configuration 5 consists of heavy R&D investment but inefficient policy coordination, imply that R&D investment (managerial behavior) has certain substitution for policy coordination (institutional condition) and firm strength (organizational condition).Lastly, both configurations 3 and 4 contain policy coordination and R&D investment conditions; however, configuration 3 consists of stronger firm strength condition while configuration 5 consists of stronger regional technology source condition, means that, enterprise strength (organizational condition) has a substitution for technological resources (resource condition).

The Configuration for Non-High TC Performance
Essential conditions that associated with non-high-performance TC in spin-off enterprises were also analyzed.Setting original consistency threshold at 0.80 and the case frequency threshold at 1, three configurations was found producing non-high performance, as shown in Table 4, namely N1, N2, and N3, which are all characterized by poor coordination of governmental policies and weak investment on R&D organizationally.Implying that when the coordination degree of local TC policy is 1 3 Journal of the Knowledge Economy (2023) 14:3453-3488 inefficient and the R&D investment is insufficient, non-high-performance TC was more likely happened, regardless of technology sources and the firm strength of TC enterprises.

Robustness Test
In QCA analysis, the raw consistency threshold selected determined the number of configurations entering into the minimization analysis, thus affecting the final results.Referred to Zhang and Du (2019), raised the consistency threshold from 0.85 to 0.90, i.e., used a more stringent threshold (analysis results are shown in Table 5).The overall consistency of solution increased to 0.901 and the overall coverage of solution decreased to 0.367.Comparison of results on both consistency threshold level of 0.85 and 0.90 revealed that, configurations 1, 2, and 3 in Table 5 are the subsets of configurations 1, 2, and 3 in Table 3, thus proved the robustness of the findings.

Discussion
In the midst of numerous studies on understanding of TC effectiveness, in this paper we sought to provide a substantive account of its high-performance paths.We focused on three perspectives of the result-the institution, organization, and technology resource articulated and instigated by the peers-and developed an integrated configuration framework for and provide an empirical test of five predictors, namely policy supply, policy coordination, technology source, firm strength, and R&D investment.We explored the configurations of these factors behind the manifestations of institution, organization, and technology resource using confirmatory factor analysis (CFA) and Fuzzy-set qualitative comparison analysis (fsQCA).
Our analysis revealed five empirically relevant configurations.When relating back to the cases in which these configurations were embedded, we identified three distinct effective TC paths.The first, dominator, pertained to the effective regional policy supply and powerful enterprise strength such like large firm size and high educated technical personnel, even if the regional technology source is deficient.The second, devotee, characterized by effective policy coordination, intensive R&D investment, and abundant technology source or powerful enterprise.The last, investor, featured by poor policy coordination and poor regional technical resources, while favorable regional policy supply and R&D investment.
Further comparison analysis reveals that managerial behavior has certain substitution for organizational and institutional conditions.Implying the government put efforts into improving regional policy supply and policy coordination, while management of TC enterprises increases R&D investment are feasible approaches in facilitating TC.Moreover, organizational conditions have certain substitutability for resource conditions.The government may seek ways to hammer at the provision of regional technical resources, for instance, increasing number of local universities to improve spin-off outcomes.Managers devote time to expanding enterprise scale, enhancing employee education background, and increasing the number of technical personnel, which are more likely accelerate TC performance.

Theoretical Contribution
Our work makes two main contributions to the literatures on technology commercialization (TC).First, by staying tuned to the notion of technology transfer is a contingent effective result (Bozeman et al., 2015;Cunningham & Paul, 2018), it helps open up the black box of the causal complex TC result that unfolds from interplay influence of institutional elements (policy supply, policy coordination), resource elements 1 3 Journal of the Knowledge Economy (2023) 14:3453-3488 (technology sources), organizational elements (enterprise strength), and managerial behavior elements (R&D investment).This helps us understand the different ways in which effective TC can be developed, as represented by the three distinct paths we identify.Along the dominator path, a supportive policy supply context that inspires innovation gives way to technology spin-off and a powerful enterprise context cultivating knowledge opportunity, which in turn gives way to TC success.Along the devotee path, a responsive policy coordination context that assists technology transfer, and intensive R&D investment as well as abundant technology source that facilitate outsourcing or utilization of cutting-edge technology, or a powerful enterprise lead way to effective TC.Along the investor path, the lack of perceived policy coordination and regional technical resources adversely influences TC result, but is backed by favorable regional policy supply and intensive R&D investment to solve knowledge deficient problems by means of new channel like outsourcing.
Our result shows the resource element has sort of substitution effect on the organizational element, so do the managerial behavior element for organizational and institutional elements.Under certain circumstances, policy supply and R&D investment are sufficient conditions to high-performance TC.These findings can help enrich our theoretical language around TC and appreciate its diversity; partially validate, expand, and deepen Bozeman's technology transfer framework.
Furthermore, our work highlights the conjunctural as well as equifinal nature of causal relationships in the generation of effective TC.Against traditional focus on fragmental importance that can simply be added to the whole explained variance, our results indicate that elements that are commonly seen as significant are actually enwind with each other and not sufficient conditions by themselves in explaining given outcomes.Actually, their significance may vary over condition or be altogether peripheral in nature.R&D investment is a particularly forceful example in this regard.While it is seen as fundamental for the innovation opportunities (Cockburn et al., 2000;Pandza & Holt, 2007), our results show that it is in some occasions peripheral in effect and a necessary condition at best.
Second, previous TC study focused a lot on European and US regions (Frondizi et al., 2019;Hou et al., 2019), while the theoretical applicability of TC systems varies from country to country (Clarysse et al., 2005;Cunningham et al., 2018).In China, the government has established well-rounded set of distinctive policy system to facilitate TC.Given the discrepancy of policy systems among government networks at multiple levels and of regional technological resources as well as its distinct humanistic background, condition combinations for TC effectiveness in the country are even more complex (Du & Zhang, 2018;Huang et al., 2013;Lu & Wang, 2012).Our work evaluates supply feature as well as coordinating feature of TC policy at different stratifications, take into consideration not only Chinese TC practices, but also Chinese science and cultural context, to derive three and four secondary conditions affecting TC policy supply and coordination respectively.Our findings helped to constitute the premise and basis for the qualitative comparative analysis, and are instrumental in better comprehending the practical contexts, consequently enhance the intrinsic effect of this empirical study.

Practical Implication
In addition to the implications for literature, our work has important implications for TC management since it suggests that, in order to be effective, managers need to dedicate efforts to move to a holistic TC system concerning the alignment of institution, resource, and organization perspectives, across its different elements.
Our work inspires TC management of different regions to fine-tune policy supply and policy coordination according to regional technology resource conditions as well as enterprise strength settings, focus on the optimization direction of condition combination of TC policy, so as to improve TC effectiveness of spin-off enterprises.
Associated with what three distinct paths signposted, management of regional TC systems will need to make constant adjustment upon different condition sceneries.On the dominator path, supportive policy supply context inspires innovation, and a powerful enterprise helps to cultivate knowledge opportunity, both give way to technology spin-off, which in turn lead way to TC success.Managers within regional TC system will have to be conscious of guaranteeing sufficient TC policy supply and pay attention to improving firm strength to build favorable environment for high-performance TC.More specifically, for spin-off enterprises' managers in supportive policy supply context, measures like increasing number of technical personnel and enhancing employee's educational background are highly recommended to ensure the strength of enterprises and consequently promote the effect of TC, even if the regional technology resource is deficient.
In addition, given the devotee path, responsive policy coordination context assists effective TC.Intensive R&D investment and powerful enterprise or alternatively abundant technology resource enable realization of successful outsourcing or utilizing of cutting-edge technology, both in turn contribute to effective TC.In this case, management of TC will must be aware of not only maintaining coordinative extant regional TC policy system, but also budgeting R&D investment or keeping firm strength to facilitate development of high-performance TC.More explicitly, for spin-off enterprises' managers in responsive policy context, maintaining intensive R&D investment helps firm achieve high-performance TC, if sustaining powerful firm or have sufficient technology source.
Finally, according to the investor path, since adverse influence of lacking perceived policy coordination and regional technology resources on TC result can be mitigated by favorable regional policy supply and intensive R&D expenditure due to solving of problem like knowledge scarce, management will have to adopt approaches such as promulgating more TC policy, spending heavier in R&D, or seeking new channels like outsourcing to generate effective TC.More precisely, for spin-off enterprises' managers in unfavorable policy coordination context, heavy expenditure on R&D or alternatively seeking connections to external technology source is extremely endorsed.

Conclusion
In conclusion, effective TC is a complex phenomenon, given the intricate joint effects of multiple perspectives and the synergistic, interactive as well as the interrelationships between conditions of institution, resource, and organization concurrently.While current demonstration of such complexity in the academic literature has been pointed to manifold factors involved or proposing integrated framework, largely within the USA and European contexts, our work takes a step towards verifying the conjunctural nature of their effects.More exactly, our work contributes by developing an integrative configuration framework and by testing it empirically in emerging economy of China with a cross-sector qualitative investigation.

Limitation and Future Lines of Investigation
There are some limitations to our research.The first concern related to the interpretability of the condition elements adopted.Although three dimensional elements were considered as institutional, resource, and organizational (Kirchberger & Pohl, 2016), it does not fully exhibit all phenomena of TC.A number of new analytical perspectives and conditional factors have been proposed in latest technology transformation studies, such as the inclusion of intelligent cluster groups in the integration research framework (Carayannis et al., 2020), essentially on account for the increasing significance of smart growth in economies.Future research should investigate in depth how smart cluster groups as well as the dynamics and interactions between and within them affect TC process, and evidence about how and why particular conditions affect TC effectiveness.
A second concern related to the measurement of technology sources.We sought to mitigate this concern through careful indicator selection, considered both the number of regional university and research institution as criterion.Nevertheless, our data did not allow us to distinguish in detail the class of universities, and the diversified technology sources, such as technologies directly outsourced from nonlocal or abroad.Future investigation may consider exploring profounder on the metrics of conditions so as to describe and analyze management practices in more detail.

R&D investment
The company leader attach great importance to R&D investment In the past three years, R&D investment in the enterprises was increased

TC performance
Number of patents (including applied)

Number of new products
Output value of new products 5. Technology resource (number of regional university and research institution see Tables 6 and 7) 1 3 2. Factors Three factors can be obtained from the above analysis: 1. Policy supply factor 1 (ZC2, TX1, TX2, TX3) ZC2: Existing TC policy often provide information, talent, venues, and other resources.
TX1: Government established scientific and technological infrastructure for strategic research is quite perfect (e.g., national laboratories).
TX2: Government established R&D institutions for generic technology is quite perfect (such as public R&D platform).
TX3: Conflicts or overlapping deficiencies of TC policies can be handled timely and properly (e.g., talent technical support, science and technology financing, and financial support).2. Policy supply factor 2 (ZX1, ZX2, TY3) ZX1: Existing TC policy tends to focus more on specific industries or industry stages.
ZX2: Existing TC policy tends to focus more on specific technologies.TY3: Existing TC policy is mainly aiming at supporting new emerging industries (such as new energy, new materials, information, and biology).

Necessity analysis
Consistency of all conditions is less than 0.9, so there is no necessary condition.(see Table 17) 3. QCA Result 1. Five configurations lead to high-performance TC The case frequency threshold was set to 1 and the consistency threshold was set to 0.85.
Through Boolean operation, five configurations of high-performance are obtained.(see Table 18) The case frequency threshold was set to 1 and the consistency threshold was set to 0.85.
Through Boolean operation, three configurations of Non-high performance are obtained.(see Table 19) Output data is as following: Fig. 1 Casual conditions for technology commercialization 3. Policy supply factor 3 (TY1, ZC1, ZC3) TY1: Research objectives of existing TC policy are usually very clear (a relatively clear description of the technical objectives and economic applications of a technology or technology area).ZC3: Existing TC policy provided sufficient fiscal support (monetary).

Funding
The study was sponsored by The National Natural Science Foundation of China(Grant No:  71772163; 42271179).Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made.The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material.If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.To view a copy of this licence, visit http:// creat iveco mmons.org/ licen ses/ by/4.0/.

Table 1
Calibration of antecedent variables and outcome variables

Table 2
1. Policy-supply-based firm strength model.Configuration 1 consists of efficient policy supply and powerful enterprise strength as core conditions and technology resource as absence condition.This configuration presents institutional and organizational elements.Efficient policy supply reflected in (i) sufficient government provision of personnel, sites, information and other resources, wellestablished strategic research infrastructures (e.g., national laboratories), wellestablished general research infrastructure (e.g., public R&D platforms), and sound construction of technology diffusion systems (e.g., national laboratories, common technology development centers, and specialized or public TC platforms); (ii) policies focused on specific industries (stages) or specific technologies, supporting emerging industries (such as new energy, new materials, information, and biology); and (iii) clear policy objectives (a clearer description of the technical objectives and economic applications of a technology or technology field), sufficient provision on funding, subsidies, tax, other financial assistance, and monetary financial support.At the same time, such conditions are accompanied by strong enterprises characteristic, as evidenced by their large size, long establishment, large number of technical staff, and high educational background of employees.

Table 3
Configurations for achieving high-performance TC

Table 10
Discriminant validity analysis result.

Table 16
Variable calibration anchors

Table 17
2. Configurations lead to Non-high-performance TC