For China, government data openness is an exotic product. Its policy development is rooted in the existing information policies and follows the general law of the development of government data openness, which is manifested as follows: China's government data openness originates from the citizen's right to know of "government information openness". At the beginning of the 21st century, with the development of information technology and the Internet, the value-added use of information resources has attracted attention from all walks of life, and the policy content has begun to emphasize "information acquisition" and "information development and utilization"; The 2015 Big Data Action Plan proposed that government departments are producers, managers, and publishers of economic, social, and livelihood data. The responsibilities and capabilities required by their multiple roles have greatly increased, and there is an urgent need to form a multi-party participation data open ecosystem. As a result, "government data openness" has gradually emerged from being embedded in policies such as e-government, national and regional informatization, and information industry. It is expected to promote the development of industries such as big data, cloud computing, and artificial intelligence through data openness to force government service reform, thus forming China's unique "government data open sharing" policy group.

The Development Stages and Characteristics of Local Government Data Openness Policies in China

（1） Stage division of local policy reference perspective

Based on the citation of national policy documents by local governments, divide the development stages of government data openness at the local level. As of June 2020, a total of effective policy document samples have been collected: 56 at the national level, 650 at 28 provincial governments, and 658 policies at 61 prefecture level cities (including vice provincial governments) (see Figure 1). Preliminary statistical findings are as follows.

From the perspective of national policy releases, from 2004 to 2012, the number of policies related to government data openness was very small, with an average of one policy being introduced every two years; More than ten policy documents were issued from 2013 to 2016, with peaks in 2013, 2015, and 2016; After 2016, the number of policy releases has remained stable at 7-8 per year.

From the perspective of local governments' citation of national policies, the trend of local policy citation is basically consistent with the development trend of the number of policies issued by the state. From 2004 to 2012, although the number of policies was small and not specific, content related to government data openness, such as government information resource disclosure, development and utilization, and inter departmental sharing, was mentioned and used as an important basis for data openness before the introduction of specific policies. For example, the "Regulations on Government Information Disclosure of the People's Republic of China" contributed 97 citation frequencies during the first peak period in 2008. From 2013 to 2016, a series of policies were issued on information consumption, "Internet plus" action, big data industry, cloud computing, artificial intelligence, network security, government services, etc. Among them, industrial development and national strategic policies have the greatest impact on local governments, making 2015 to 2016 the second peak of local reference to relevant national policy documents. After 2016, based on the management measures for sharing government information resources in China, a number of policies were introduced in the fields of government reform and government services, which expanded to support the integration of government data internal sharing and external openness, website construction, and data resource management. In 2017, there was a third peak in local references to relevant national policy documents.

Figure 1 Distribution of Text Volume and Local Government Citation Frequency of the Central Government's Policy on Open Government Data from 2004 to 2020

Based on the frequency and content analysis of local government citations above, it can be divided into the following development stages.

(1) The preparation stage of government data openness led by information disclosure (from 2004 to early 2013).

(2) The exploration stage of government data openness under the guidance of national strategy (from the first half of 2013 to early 2016).

(3) The rapid development stage of government data openness and sharing driven by government reform (from the first half of 2016 to present).

（2） Characteristics of Stage Development from the Perspective of Policy Tool Analysis

Firstly, from the perspective of national policy themes and contents related to government data openness (see Figure 2), the period from 2004 to 2012 was a policy development stage with information disclosure and information resource development and utilization as the main themes. Government data openness has not yet been officially recognized, but the development and socialization of government information resources have been repeatedly mentioned in policies, which can be regarded as the predecessor and foundation of China's government data openness. During this period, the habits of selecting government administrative norms and policy tools in the information environment are forming; Since 2013, plans for the development of the information industry such as information consumption and "Big Cloud Intelligent Connection" have been successively released, and the open sharing of government information resources and government data has been concentrated in the safeguard measures of policy documents; Since 2016, management measures and work plans for government information resource sharing, directory compilation, system integration, and other management methods closely related to government data openness have been successively introduced, marking the gradual shift of China's government data openness policy from passive to active, from behind the scenes to the front desk, and from the information age to the data age.

Secondly, in terms of the distribution of policy tool types, there is a high degree of similarity in the usage preferences of countries and regions, arranged in the order of "command type>capacity building type>incentive type>persuasion type>learning type". The use of command type and capacity building type tools is dominant (both exceeding 1/3), and the proportion of incentive type tools remains stable at 10% to 15%, while the proportion of persuasion type and learning type tools is relatively small, staying at around 5%. The extensive use of command based tools reflects the distinct "top-down" design characteristics of China's government data openness policy implementation, especially in the early stages of construction (the first stage) and the period of government reform (the third stage). The most feasible and authoritative approach for local governments is to promote local government data openness by referring to national policies. The same situation is also reflected in the stable use of incentive tools, such as the national use of incentive tools such as "demonstration pilot" and "increasing funding budget" to promote the implementation of local data openness policies. Of course, local governments also have a certain degree of autonomy in the use of policy tools, mainly manifested in the fact that the use of capacity building tools at the local level is generally higher than that at the national level. The reason for this is that compared to command based tools, which require regulations, plans, opinions, methods, etc. with higher effectiveness levels to be promulgated and implemented, capacity building tools have the characteristics of "appropriately relaxed effectiveness levels, stronger flexibility, and promotion in the form of work plans, action plans, etc.", which are more suitable for local governments to provide information, talent, technology, and other support needed for government data openness, as well as to formulate supporting work systems and assessment supervision. The use of persuasive tools decreases with the decline of administrative levels, and decreases with the expansion of the dissemination of data openness concepts and the increase in the use of substantive policy tools, especially in the third stage where their proportion is less than 5%. The development trend of the application of learning tools is exactly opposite to that of persuasion tools, gradually forming a combination of command and capacity building tools at the local level, such as the coordinated use of expert systems and conference events.

Figure 2: The main goals, policy themes, and distribution of policy tools for the three-stage development of government data openness in China

Thirdly, looking at the distribution of policy tools from the perspective of open data, utilization of data, public value, and external environment as the main links of the government data open ecosystem, it can be found that the first stage is characterized by the "hardware" preparation of infrastructure construction, awareness and institutional preparation of information sharing, the formation of work systems and evaluation mechanisms for information release, and the formation of awareness of providing "open utilization" to the outside world; The second stage is characterized by the emergence of the concept of "data governance", incentive policies to promote industrial development and public services, and local governments focusing more on improving relevant systems; The third stage is the stage where the focus of construction returns to the internal factors and environmental conditions of government departments, such as infrastructure, institutional construction, and working mechanisms, from the national to the local level. The main characteristics are the trend of "data governance" becoming more detailed and legal, the increasing willingness for external cooperation, further exploration of economic and public values, and the increasing differentiation in the use of local tools.

The concept of "institutional embeddedness" in new institutionalism holds that policy tools are not a single abstract project, but are influenced by factors such as institutional environment and policy subject concepts. After the implementation of government data openness policies at each stage, some problems and contradictions in the previous stage can be alleviated, and certain institutional habits and promotion paths can be formed. However, at the same time, new problems will arise. Therefore, the weak links after the implementation of policy tools in the previous stage can serve as the analysis results of policy implementation in the current stage, as well as the background of policy tools for policy implementation in the next stage.

To further understand the application of government data openness policy tools in the rapid development stage, and to provide reference for local governments at different levels to promote the implementation of data openness policies in the context of the big data era, the following will focus on analyzing the characteristics and differences in the application of policy tools in the third stage.

Characteristics of Policy Tool Application in the Rapid Development Stage of Government Data Openness in China

（1） Two dimensional analysis of the application of policy tools at the national level

The rapid development stage of government data opening in China is mainly marked by entering the "13th Five Year Plan" stage. A series of policies such as national informatization, big data, "Internet plus government services" are popping up. The government information resources are defined at the national level. A total of 40 policy documents related to the opening and sharing of government data have been issued (see Table 1). They are mainly normative documents such as the State Council's planning, guidance and implementation plans, as well as departmental work documents such as action plans and work notices issued by the National Development and Reform Commission, the Office of Cyberspace Affairs, the State Market Supervision and Administration, the Ministry of Transport, the Ministry of Industry and Information Technology, the China National Intellectual Property Administration, etc. The average strength of the effectiveness level is 2.3, and the standard deviation is 0.94. The policy strength is quite different. The distribution of effectiveness levels and policy intensity is very similar to the second stage (average 2.3, standard deviation 0.92), and the release of government data openness related policies tends to be stable.

表1 政府数据开放领域相关的国家政策

|Excel下载

Table 1 National Policies Related to Government Data Openness

The national policy themes in the rapid development stage include: national informatization strategy, Internet plus government service, government information resource sharing and government information system construction, government openness, reform of "decentralization, management and service", development of strategic emerging industries, big data application to improve business environment and promote innovation and entrepreneurship, information disclosure in specific fields, network security, etc. 40 policies and 88 policy tool units in this stage are included in the analysis list of policy tools in the government data open ecosystem of this study (see Table 2). The application of policy tools is as follows.

(1) Basic distribution: Compared with the first and second stages, the distribution is closer to that of the first stage. Command based tools have an absolute advantage (47.73%, 50% in stage I, flat), followed by capacity building tools (30.68%, 21.9% in stage I, ↑ 8.78%), motivational tools (15.91%, 12.5% in stage I, ↑ 3.41%), learning tools (3.41%, 3.1% in stage I, flat), and persuasive tools (2.27%, 12.5% in stage I, ↓ 10.23%), with persuasive tools showing a significant decline in the third stage. This is related to the use of persuasive tools at the national level, which often rely on "propaganda and promotion" and "encouragement and call" tools to encourage public socialization and social forces to enter the fields of information resource development and big data application development. The second and third stages of incentive tools replace the functions of persuasive tools in terms of special investment in funds, projects, and technology.

(2) Command based tools (47.73%) have shifted from primarily providing guidance and authoritative work systems in the first and second stages to emphasizing the use of guidance, planning, authoritative work systems, and system building tools. The guidance and instruction tools are gradually being refined and applied to the description, quality, security, sharing, and information release of data in the "open data" process. The concept of "data governance" that began to emerge in the second stage is further integrated, and clear requirements are proposed at the national level. For example, in terms of data description, the "Guidelines for Compilation of Government Information Resource Catalog (Trial)" and the requirement to "formulate a government data sharing open catalog" are introduced to clarify the current status of data resources internally and clarify the content of open data externally; In terms of data quality, it is required to ensure the integrity, availability, and timeliness of data in accordance with the principle of 'who is in charge, who provides, and who is responsible'; In terms of data security, it is required to implement network security protection measures for system integration to ensure the security of the sharing platform; In terms of data sharing, it is required that all departments accelerate the construction of a standard system for government information sharing; In terms of information dissemination, provide guidelines for the development of government websites and guide government departments that have not yet launched specialized platforms for government data openness to release relevant data information. The planning tools are mainly used in data infrastructure and social utilization, requiring the acceleration of the construction of a national unified open platform, "big data platform", "data center", and "data sharing big platform" for government data. In terms of social utilization, plans for vertical data information opening and sharing have been introduced, such as the sharing and opening of transportation, market supervision, national intellectual property patents, etc. The work system (authority) is mainly aimed at organizing and implementing cross regional and cross departmental sharing of government information resources. For example, it requires "local governments at or above the city level" to clarify the competent department and establish a "joint meeting responsible for sharing and coordination". Each department is required to "report the sharing situation to the joint meeting at the end of February every year"; To achieve mutual recognition and sharing across regions, levels, and departments, the State Council needs to open real-time data interfaces for business systems to online government service platforms in various provinces (regions, cities) as needed. The system building tools are used in Internet plus government services, data security, data resource management and other links, including the construction of business support system, platform system, key technology support system, evaluation system, etc. of government services, the construction of network security standard system (including national and industrial standards for network security management, products, services and operation security), and the construction of government data resource system to ensure "one number one source, multi source verification, and dynamic update".

表2 快速发展阶段国家层面基于政府数据开放生态系统的政策工具运用情况分析

|Excel下载

Table 2 Analysis of the Application of Policy Tools Based on Government Data Open Ecosystem at the National Level during the Rapid Development Stage

(3) Compared to the first stage (21.9%) and the second stage (20%), the proportion of capacity building tools used is higher (30.68%), with a particular emphasis on the use of "data security" work systems (organizations) and talent support tools. In the construction of big data platforms and the sharing and utilization of government information resources, propose a classification and grading management system for data resources, a network security guarantee mechanism, a user information protection system, etc., and study and explore the "classification and grading system for government information resources" and "data security management methods". For the cultivation of talents in network information and data security, the country has expressed support for enterprises, universities, and vocational schools to carry out education and training related to network security.

(4) Compared to the second stage (24%), the proportion of incentive tools (15.91%) has significantly decreased, only slightly higher than the first stage (12.5%). Some of these tools continue to be used in the second stage, such as the pilot reform of public data opening and utilization, demonstration and selection of major big data development projects, and punishment and rectification for non compliant sharing of government information resources. Additionally, the scope of priority projects and policy pilot programs has been expanded. For example, platform projects that provide basic support for data sharing and openness, public welfare projects that provide public services, and policy support and guarantee projects that establish a market circulation mechanism for data elements are given "priority" in major big data engineering projects. In addition, new pilot projects for government transparency, standardization and normalization of grassroots government transparency (100 counties and cities), and pilot projects for public information resource openness will be added. Compared to the second stage, the third stage focuses on the data openness aspect, especially in the comprehensive promotion of government services and pilot sharing and utilization of information resources within the government.

(5) The proportion of persuasive tools (2.27%) and learning tools (3.41%) has further decreased, focusing on data security and social utilization, targeting government workers and non-specific members of the public, similar to the first stage. In terms of data security, we encourage all sectors of society to "develop network information security protection and utilization technologies", cultivate network security awareness, and propose that "governments at all levels organize regular network security publicity and education". Public participation tools are mainly used in areas that require public participation and supervision, such as market regulation, economic and social development, implementation of measures to benefit the people, and complaints about network information security. It is required to strengthen participation and timely handle feedback.

(6) In the rapid development stage of the government data open ecosystem, the focus of national level construction is on information resource management, data sharing and security, government information resource sharing and utilization, as well as continuing the support of the second stage big data application project in public services and social governance. This is reflected in more than 75% of policy tools being applied in the two major links of open data and data utilization. Command based and capacity building tools are distributed in a balanced manner in two aspects, serving respectively in infrastructure construction, data resource management, data security and sharing, information dissemination and social utilization, and industry and market cultivation. The target audience of persuasive and learning tools has shifted from government officials to non-specific members of the public, emphasizing public participation and supervision in decision-making related to cybersecurity and people's livelihoods.

（2） Two dimensional analysis of the application of policy tools at the local level

In the rapid development stage, policy documents related to local governments (Class A and Class B) and government data openness are mainly distributed in policy documents of government openness and service platform construction, government information system integration and sharing, e-government and digital government construction, public information/data resource management and sharing openness, big data application and industrial development, industry big data management and utilization, specific field information disclosure (major engineering projects, public resource allocation, social welfare undertakings), credit information collection and utilization, and other subjects. Among them, the most cited national policies are the "Notice of the State Council on Issuing the Interim Measures for the Management of Government Information Resource Sharing" (2016), the "Cybersecurity Law of the People's Republic of China" (2017), the "Notice of the General Office of the State Council on Issuing the Development Guidelines for Government Websites" (2017), the "Guidelines for the Compilation of Government Information Resource Catalog (Trial)" (2017), and the "Notice of the General Office of the State Council on Issuing the Implementation Plan for the Integration and Sharing of Government Information Systems" (2017). After experiencing two stages of transition from information disclosure to government data openness, from encouraging social utilization to effectively stimulating the development of big data cloud computing industry and promoting the development of government data openness, local governments will return the focus of construction to within government departments, and make efforts in internal factors and environmental conditions such as infrastructure, institutional construction, and work mechanisms.

From the 453 policies collected from 28 regions in the rapid development stage of Class A governments in this study, it can be seen that the main body of policy promulgation at the local level is the provincial/municipal government, followed by the local Development and Reform Commission, Economic and Information Commission, Information Technology/Big Data Work Leadership/Promotion Group, Big Data Development Management Bureau/Industrial Development Leadership Group. In addition, a small number of policies are issued by departments such as agriculture, transportation, market supervision, auditing, and public security. The average effectiveness level is 2.3, which is slightly lower than the average intensity of provincial-level urban policies in the first and second stages (around 2.7). Implementation opinions, implementation plans, work plans, work notices, and notifications are the most common, followed by management measures and development plans, with a standard deviation of 0.93. The level difference is the same as the national policy difference in this stage (0.92); Compared to the 455 policies collected from 61 B-class cities during the same period, the average effectiveness level is 2.36. Work notices and notifications with lower effectiveness levels are the most common, followed by implementation plans, work points, and action plans. A small number of regulations with higher effectiveness levels have been introduced, with a standard deviation of 0.90. The level difference is basically the same as that of A-class governments (0.93) and second stage governments of the same level (0.98) during the same period. It can be seen that in the rapid development stage, local government data openness policies, under the requirements of national strategy and the deployment of the State Council, have further implemented important tasks such as platform construction, system integration, and resource construction for government data openness and sharing. On the one hand, policy documents for operational reference have been issued in the form of implementation opinions, implementation plans, and work plans. On the other hand, before the national higher-level laws and regulations on government data openness are introduced, local governments have taken the lead in exploring special policies in the fields of public data and government information resources. From 2017 to 2019, A-class governments such as Zhejiang, Shanghai, Beijing, Sichuan, and Jilin, as well as B-class governments such as Chengdu, Fuzhou, Ningbo, and Siping, successively issued local normative documents such as the "XX Province/City Public Data/Government Data Resources and One Stop Service/E-Government Management Measures" and the "XX Province/City Government Data Resource Sharing Management/Sharing Open Management Measures", which have become important basis for local government data openness.

From the distribution ratio of policy tools (see Figure 3), there is a slight difference in the ranking of policy tool types between local governments and the national level, which are "command oriented>capacity building oriented>incentive oriented>learning oriented>persuasion oriented", The ranking and proportion of policy tool types for Class A and Class B local governments show a high degree of similarity. Unlike the second stage, the use of capacity building tools is greater than the use of command based tools. For the first time, the use of capacity building tools has surpassed command based tools to become the key tool type used by local governments. The reason is that government reform actions such as government openness, government services, system integration of government information resource sharing, and resource construction are conveyed from the central to the local level. At the national level, the use of command based tools and capacity building tools for open data links accounts for 70%. Clear instructions have been made on principles, regulations, systems, and work mechanisms for infrastructure construction, data resource management, data sharing, data quality, data security, and information release, as well as platform construction and security. System construction in terms of protection Technical and talent support have proposed planning schemes or guidance opinions, which need to be implemented by local governments in specific work. However, it was not until the third stage of government data openness and sharing that important concepts and main responsibilities such as public data, open data, and government information resources were gradually clarified. In this situation, the effectiveness of policy implementation largely depends on the reform results within the government. The role of capacity building tools lies in transforming the requirements of command based tools for government departments and staff into practical capabilities for government data openness and sharing and supporting the improvement of public services.

Figure 3 Comparison of the distribution of policy tool types between national and local governments during the rapid development stage

（3） Differences in the use of policy tools between national and local governments

From the perspective of the role of policy tools in the government data open ecosystem during the rapid development stage (see Figure 4), the overall structural distribution between the national and local levels is similar, with the difference being that public value at the local level is greater than the external environment. The distribution pattern and reasons are analyzed as follows: ① Special management measures have been introduced and a large number of government data open platforms have been launched, providing policy and platform support for the realization of public value. The "open" phase, due to the focus of construction returning to within the government during the rapid development stage, further strengthens the management of data resources and the construction of data infrastructure related to government data openness. This is mainly reflected in achieving the goal of integrating and sharing government information resources and system transformation, as well as establishing a matching organizational structure, working mechanism, resource system, standard system, security system, etc; Moreover, the government provides project funding support and technical guidance for the construction of local "open" links through incentive and command based tools such as reform pilot programs and development guidelines. In practice, the construction of local government data open platforms has grown at a double rate from 2018 to 2019, with a total of 142 projects as of the second half of 2020 The "utilization" link (20.45%, 15.87%, 19.62%) has significantly decreased compared with the second phase (40.00%, 30.44%, 32.11%), because the second phase introduced a large number of national strategic plans and local action plans in terms of big data, cloud computing, and "Internet plus". By the third phase, the national level mainly introduced supporting policies in terms of project construction and application, and the number and frequency of release have declined, while the local level is more "reference makers" of the "Big Cloud Smart Connect" industrial strategy and action plan, and the richness of the use of policy tools needs to be improved. ③ In terms of public value, the proportion from the state to the local (9.10%, 15.27%, 12.22%) has increased compared with the second stage (8.00%, 7.82%, 5.82%), which is mainly due to the introduction of Internet plus government services and government affairs openness and other related policies, reiterating the importance of "mutual recognition and sharing of government information resources, multi use" and promoting "the joint participation of the government, the public, and enterprises", requiring the construction of a "new pattern of government services" and the creation of a "responsive and open" government service platform. ④ The proportion of national and local links in the "external environment" stage is similar to that of the first stage (37.5%, 11.76%, 4.02%), with a significantly higher proportion at the national level than at the local level. Due to the country's renewed emphasis on the importance of "sharing" and "security", especially in terms of organizational implementation and protection of related rights, there are new institutional arrangements. The establishment of the "Leading Group for Government Openness" and coordination mechanism, as well as the evaluation system for government services, the revision of the Information Disclosure Regulations once again clarifies the system of "the office (room) of the department implementing vertical leadership is in charge of the government information disclosure work of this system" (authority), and requires the establishment of a sound system for protecting government website user information in related rights protection.

In terms of the application of government data openness policy tools, it has gone through the preparation work of the "openness" and "environment" stages in the first stage, the extensive use of incentive tools in the "utilization" stage in the second stage, and the distribution of policy tools in various stages in the third stage has reached a balance, with slight changes at the national level, while the distribution at the local level is highly similar to that in the second stage, and the distribution ratio between Class A and Class B cities is also very similar, with only new changes in the use of tools in some stages (see Table 3).

Figure 4 Comparison of the distribution of policy tools between national and local governments during the rapid development stage

表3 快速发展阶段省级和副省级/地级政府的政策工具运用在政府数据开放生态各环节的分布情况

|Excel下载

Table 3 Distribution of Policy Tools Applied by Provincial and Sub Provincial/Prefecture level Governments in Various Links of Government Data Open Ecology during the Rapid Development Stage

(1) Preference for the use of key policy tools in command based tools.

A-class and B-class governments will focus on using regulatory standards, management methods, and system building tools in the third stage to achieve the important content of "openness" and "public value" in the second stage guidance and instruction tools, such as clarifying information resources and business boundaries, building shared platforms, achieving system integration, and enhancing the functions of public service platforms Regulatory standard tools are used in the management of data resources such as collection, description, sharing, security, openness, and utilization of government data. They prioritize the development of standards for data exchange, sharing, and open security to ensure the safe operation of government information resource sharing, openness, and services. As a result, each province/city has formulated regulations and standards for government information collection, directory standards for open datasets of public information resources, standards for government information/data sharing, as well as comprehensive technical and business standards covering "government information resource data collection, data quality, directory classification and management, shared exchange interfaces, shared exchange services, multi-level sharing platform docking, platform operation management, network security protection, etc. The above standards and specifications are mostly scattered in government information resource sharing, government transparency standardization and other work documents in most cities. Guiyang City has specifically issued the "Implementation Plan for the Construction of Guiyang Big Data Standards" (2017), which is in line with its urban strategy for the development of the big data industry. It involves key common standards for public data such as "government data classification, openness, sharing, and security management", group standards and enterprise standards for "data sharing, data rights confirmation, data pricing, data circulation, product transaction interface specifications" and other "data utilization" links that affect the development of the big data industry and market, as well as "big data technology standards", which are all included in its big data standard construction scope The management method tool is applied to the management of government data resources and the standardization of open utilization of specific data types. The former applies to government departments, while the latter applies to government departments, as well as institutions and enterprises that collect and provide relevant data services. The management measures introduced by Class A and Class B governments include the sharing and exchange of government information resources, the construction of government websites and portals, the construction of government information systems and projects, the management of government data resources and security, and the construction of common infrastructure and resources related to data "openness", as well as the management of data openness in vertical fields and specific industries such as geographic information data, transportation data, health and medical data, and credit information collection and application. In addition, Class A governments have also introduced management measures for public data and One Network Office, e-government and government data management, public data and e-government management, to prepare for giving play to the open sharing of data to improve social governance and public service capabilities, while Class B governments have further standardized the industrial application of big data in urban governance by going deep into online taxi information security, Internet hospital data security and confidentiality, urban renewal of basic data, etc. ③ The system construction tool is mainly used within government departments, involving public services and data security, that is, the data sharing and exchange platform support system and technical system for government services, as well as the construction of data resource management, data sharing, and information system security guarantee system. The application of this tool has a certain degree of cross correlation with regulatory standard tools. For example, the security support system needs to determine the security level attributes of data based on the classification and grading standards of government data, and the management and service of government data resources need to develop government information resource directory system standards and sharing standards; In addition, B-class governments have proposed to establish an efficiency evaluation system for government services, which is related to the frequent use of assessment and supervision tools by B-class governments.

(2) Among incentive tools, Class A and Class B governments tend to use similar types and functions, as evidenced by the widespread use of punishment and rectification tools in various aspects of the government data open ecosystem, a significant increase in the use of tools in public services, and differences in the use of Class A and Class B government tools mainly reflected in policy pilots, demonstration evaluations, licensing, recognition and rewards.

① The application of punishment and disciplinary rectification tools is the widest and most numerous, with the smallest difference in the use of A-class and B-class government tools. They are used to ensure the implementation of government information resource sharing management methods, public data openness management methods, data security management methods, etc. For data sharing, data resource management, data infrastructure, and data security links that involve failure to construct and operate platform information systems in accordance with regulations, incomplete system construction, failure to fulfill work coordination leadership responsibilities, and the existence of safety accident responsibility issues, they will be handled according to the basic process of "superior orders for rectification - failure to rectify within the prescribed time limit, notification and evaluation - responsible persons will be punished according to law/criminal responsibility will be pursued according to law".

② Demonstration selection, policy pilot, government procurement and outsourcing tools have been applied by local governments in the public service sector. The demonstration selection tool is applied to urban governance and public service (cooperation), collecting big data products and application solutions from society and giving rewards, mainly involving government data sharing and opening, data trading and circulation, as well as big data application and service innovation in macroeconomic, fiscal and taxation, public safety, environmental protection, food, medical, education, tourism and other fields. The policy pilot tools place more emphasis on the value mining of data openness and sharing in the public domain, manifested in the pilot of information benefiting the people, the innovation pilot of the "One Number, One Window, One Network" government service model, the pilot of public information resource openness (including national designated and local implementation), and the pilot of standardization and normalization of government openness, emphasizing the improvement of service efficiency, simplification of work processes, smooth service channels, innovation of service methods, and strengthening government response based on government data sharing and opening. Government procurement and outsourcing tools are used in the construction of data infrastructure and public services, such as building data centers to ensure the sharing and secure utilization of government big data, purchasing government cloud platforms to improve e-government efficiency and public service quality, and providing services through purchasing services and agreement agreements by enterprises.

③ Financial budget and financing cooperation tools are used in data resource management, infrastructure construction related to data sharing, and big data industry incubation. This is reflected in the inclusion of construction funds for promoting the integration and sharing of government information systems into government fixed assets investment, the inclusion of government information resource sharing catalog preparation, platform operation and maintenance and other related work funds into the department budget, the provision of sufficient funds for the construction, operation and maintenance of public basic platforms, the gradual reduction of "information island" systems and private networks with low utilization or even the cancellation of their operation and maintenance funds, so as to solve the "island" problem mentioned in the second phase. Financing cooperation tools promote the development of the big data industry through government funding, attracting social capital, establishing a big data industry development guidance fund, and expanding financing channels for big data enterprises.

④ Some B-class governments have innovated in the use of recognition, reward, and licensing tools, manifested in the use of tools in public service (cooperation), data sharing and management, and big data industry incubation. For example, in Jiamusi, Yinchuan, Shangqiu, Fuzhou, and Qiannan prefectures, while using incentive tools for punishment and rectification, they also recognize and reward units or individuals who have completed work tasks ahead of schedule, achieved high comprehensive evaluation and good practical results, and outstanding achievements in shared management work in government service work; Fuzhou, Guiyang, and Xiamen also commend or reward citizens, legal persons, or other organizations that have outstanding performance in researching, analyzing, and mining government data to promote the development and innovative application of the big data industry. In terms of the use of licensing tools, Class B is slightly more common than Class A. Guiyang, Huaibei, Fuzhou, Xi'an and other cities have signed agreements or management measures to clarify that government data can be authorized to third-party companies for development or service operation, such as Guizhou's Block Data City Construction Co., Ltd. and Huaibei's Construction Investment Commercial Big Data Information Co., Ltd. Hainan has introduced and cultivated a number of leading big data enterprises through government data openness and authorized operations.

(3) Local governments generally attach great importance to capacity building tools, and their distribution in the four major aspects of openness, utilization, public value, and external environment is basically consistent with that of command tools. The most commonly used work system (organization) and assessment and supervision tools, followed by talent support tools. A-class governments have a slightly higher proportion of tool utilization in public value and external environmental aspects than B-class governments, and have more advantages in the use of technology support tools.

① In terms of work system (organizational) tools, the first and second stages have already formed some work mechanisms and institutional norms in data sharing and data security. The third stage is the improvement of existing institutional content and innovative exploration in data quality, data description, organizational implementation guarantee and other aspects. In terms of data sharing, based on the completion of the sharing and exchange platform construction, various regions will establish a "unified acceptance, platform authorization" data sharing authorization mechanism, a sharing application mechanism, and a deadline feedback mechanism, emphasizing sharing security and efficiency; In terms of the shared list system, there have been innovations based on the existing resource list, directory list, responsibility list, and power list, such as the negative list system and its review mechanism (Hainan), which will be discussed and reviewed by the provincial information technology department in conjunction with the legal department and the compilation department. After public announcement, government information resources outside the negative list will be unconditionally shared. In terms of data security, on the basis of establishing a general classification and grading management system, confidentiality review system, risk assessment and security protection, and disaster recovery mechanism, we attach great importance to the improvement of the management system and security training mechanism for security management personnel. In terms of data quality, Class A cities have taken the lead in exploring and proposed the principle of "synchronous collection and real-time update" based on the principle of "who is in charge, who provides, and who is responsible", to ensure that data is accurate, complete, timely, and available. Public information resource providers are responsible (Hainan, Hubei, Jiangxi), or the principles of "who collects, who is responsible" and "who verifies, who is responsible" are proposed, with public management and service agencies and municipal departments responsible for quality responsibility, and the city's big data center responsible for quality supervision and real-time monitoring and evaluation (Shanghai). In terms of organizational implementation guarantee, Class A cities have established a job responsibility system for public information resource management and sharing work, requiring each unit to determine the responsible department and specific responsibilities. Innovation has also been made in the design of similar systems, such as the establishment of the "cloud chief system" and "data specialist" system, responsible for promoting the construction of digital cities and data resource management and sharing openness (Guizhou).

② Assessment and supervision tools are widely used at the local level, especially in the public service sector, including the assessment of "Internet plus+government service", government affairs openness, government website data information release, and public data openness, and the assessment of the reform effect of government services in combination with public complaints, negative comments, media supervision, expert review, third-party evaluation and other methods. Assessment and supervision tools are the most commonly used tools before the formation of a series of comprehensive work systems, often achieved through quantitative evaluation of work tasks to achieve internal performance management goals.

③ Technological support tools are continuously applied in data security, data sharing, data infrastructure construction, and public service sectors. The application of technology support tools is particularly important in the public service sector, which is the guarantee for achieving government data sharing, platform interconnection, and integrated applications. Specifically, it supports the construction of integrated government service websites and platforms. Some A-class pilot cities are connected to the national government service platform, and most A-class governments provide one-stop government services on computers and mobile phones. Platform integration applications are carried out in online service halls, government communication apps, government portal websites, 12345 convenient service platforms and other fields to meet the needs of the public for querying data, using data, and handling administrative services. In terms of data security, we will build a government data aggregation and sharing security control platform, a government external network security access platform, and a government app security detection platform, improve monitoring and prevention systems, and achieve the identification and protection of new methods and technologies that threaten network security. In terms of data sharing, Class A governments further require the achievement of the goal of "cross departmental, cross regional, and cross hierarchical" government information resource sharing. Through the transformation and integration of government systems, third-party information technology companies can connect with local and departmental governments to complete data migration and platform conversion. This is one of the ways to solve the problem of "stock" data sharing and eliminate "islands". In terms of infrastructure construction, data center projects such as big data centers, big data bases, and big data resource centers have become the main carriers of technical support. In the third phase, they will be fully deployed as data storage and disaster recovery bases, gathering public, enterprise, and government data, and providing centralized basic environment and computing, network, storage, software and other resources and services. The development of B-class governments is slightly slower, and they are still focusing on the construction of data sharing platforms, government systems and cloud migration, basic data sharing libraries, and mobile platforms for government services.

④ The frequency of using talent support tools has increased, shifting the focus from the second stage on industrial incubation (supporting big data talent cultivation) to emphasizing data sharing, information release, public services, and industrial incubation. The target audience is government department staff, and business training related to data openness, sharing, and public services will be included in the cadre training curriculum to enhance professional competence. The focus of talent support for industrial incubation is mainly on talent cultivation and introduction, such as incorporating big data cloud computing talents into talent cultivation plans in various provinces and cities, encouraging enterprises to cooperate with training institutions to carry out professional training and vocational education, and encouraging local universities to establish key laboratories and research institutes to attract and cultivate a group of professional talents.

(4) In terms of persuasion tools, Category A pays more attention to the use of public value links, while Category B pays more attention to the use of external environment links. This is reflected in that Category A cities use more publicity and promotion tools in public services, emphasize "one number, one window, one network" government services, increase publicity, and use news media, new Internet media and other real-time communication features to improve social awareness and recognition. B-class cities are more commonly used in the cultivation of digital culture, organizing academic conferences, development forums, salon lectures and other promotional and training activities, creating information and big data experience centers and achievement display windows, etc., to enhance citizens' perception of information and big data construction.

(5) In terms of learning tools, Class A emphasizes more on diversified participation in the development of public value, while Class B emphasizes more on the use of public participation and expert systems in the open and utilization process. This is manifested in Class A governments encouraging data utilization entities to cooperate with government departments' informationization, big data, and data openness entities, jointly utilizing and opening up the value of public data to form various achievements for application in social services, market supervision, administrative supervision, and other fields. The public participation of B-class government refers more to the participation in government response and feedback on information release and government transparency. Interaction is achieved through functions such as message comments, online interviews, survey solicitation, consultation and complaints, instant messaging, etc. Government departments respond and guide public opinion through transparent progress, image interpretation, and timely release. In addition, expert system tools are gradually penetrating into various aspects of the government data open ecosystem, such as the legality and security issues of data sharing, and expert discussion meetings should be held for those with strong professionalism; The construction of a digital government requires the establishment of an expert group to provide suggestions and recommendations; The policy measures, standards and specifications, project review, application collection, system construction, etc. related to public data management should be reviewed by an expert committee; Government data openness involves professional issues such as technology, law, security, and intellectual property, and should be subject to professional opinions from the Big Data Expert Advisory Committee; Before local governments introduce draft laws and regulations related to government data openness and sharing, it is necessary for a data openness and sharing expert group to attend legislative hearings.

（4） Policy tool preferences and weak links of government data openness in the rapid development stage

The rapid development stage of government data opening is also the period of China's "13th Five Year Plan". The reform of government services has become one of the most important backgrounds in this stage. Policy documents with the theme of government information resource sharing and opening pilot, Internet plus government services, and government openness have been successively issued and implemented at the national and local levels. In terms of data resource management, data collection/collection, data quality, data security, social use, and public services, the variety of policy tools has become increasingly rich. From the internal infrastructure construction focusing on the "opening" link in the first stage, to the external value extension of the "use" link in the second stage, to the internal management and capacity building focusing on the "opening" link in the third stage, to better promote the "use" "and the promotion of" public value ".

(1) In terms of the application of policy tools in the "open data" phase, there has been an increase in the use of more targeted and actionable management methods, regulatory standards, and tools, as well as a new focus on data quality. At the national level, a series of management measures, guidance opinions, implementation rules, work plans, etc. have been introduced for government information resource sharing, government services, and government transparency, concretizing and detailing the concept of "data governance". The data description, data collection, data quality, data sharing, data security, and information data release of government data openness can find policy basis at the national level. On the premise of formulating the "Directory of Government Information Resources" and "Open Directory of Government Data Sharing" required by the state in the form of guidance and instruction tools, ensuring the "integrity, availability, and timeliness" of data, safeguarding the security of sharing platforms, accelerating the construction of a standard system for government information sharing, and completing the work tasks of disclosing government department data information according to website development guidelines, local governments have focused on using regulatory standards, management methods, and system construction tools to introduce a series of standards and specifications for government information resource data collection, quality management, directory classification, sharing and exchange, platform docking and operation, network security, and given priority to formulating standards and specifications for data collection/aggregation and data quality management. Moreover, we attach great importance to the use of capacity building tools in information release, data sharing, and data security, and innovate in work mechanism tools, such as introducing a response mechanism in the performance evaluation of government transparency, exploring innovation in government information resource sharing and data openness systems based on the original classification management and list system, such as distinguishing shared information classification according to basic and business (Guangzhou), and developing a shared negative list system and negative list review mechanism (Hainan). It can be seen that the standard system construction and institutional innovation of local government data openness have been reflected in the third stage.

(2) The use of incentive and learning tools in the "utilization" stage reflects the willingness of local governments to cooperate externally: in the industrial incubation stage, financing cooperation and licensing tools are used to encourage the big data cloud computing industry to put into operation and build data centers locally, authorize and license third-party enterprises to develop and utilize government data and provide government services, website operation and maintenance services, etc; Supporting the use of learning oriented diversified participation and expert system tools, encouraging data utilization subjects to cooperate with government departments' informationization, big data, and data openness subjects, and encouraging data experts and scholars to participate in the formulation of relevant policy measures, legislative standard research, and project evaluation related to legality, compliance, security, etc. in the government data openness ecosystem.

(3) In terms of public services with "public value", the national level emphasizes system construction, while the local level extensively uses incentive and capacity building tools, such as demonstration selection, policy pilot, government procurement and outsourcing, recognition and rewards, etc. Corresponding to the "One Number, One Window, One Network" government service reform pilot and demonstration selection, government procurement and outsourcing of big data products related to public service applications, individuals and units who have made outstanding contributions to promoting government data openness and sharing and improving public service quality will be commended and rewarded. In addition, the capacity building tool combines examination and supervision with technological support. The former evaluates government services, government openness, and data sharing through public feedback, media supervision, expert evaluation, and third-party assessment; The latter supports the construction of integrated government service websites and platforms, as well as the integration of multiple platforms such as online halls, government apps, portal websites, and convenience platforms. In addition, local governments also use persuasive tools, such as A-class cities using news media, new media, etc. to promote the "One Number, One Window, One Network" government services, and B-class governments promoting the formation of digital culture through lectures, salons, training activities, big data experience centers, and achievement display windows.

From this, it can be seen that in the rapid development stage of government data openness, in addition to implementing various tasks related to government service reform, the concept of "data governance" is deeply integrated into all aspects of the government data openness ecosystem. Professionalization, refinement, and local characteristics are forming. The policy tool content of the third stage of government data openness and sharing partially responds to and solves the problems of information silos, low data quality, and lack of digital culture in the first and second stages. The construction of various aspects of the local government data openness ecosystem is also taking shape, but the implementation path and ideas of data openness still need to be organized, and may face the following challenges in the future.

(1) Problem to be solved. At present, the concept of government information resource development and utilization has gradually been replaced by government/government/public data openness and utilization. Information openness and government data openness have been included in the scope of government openness in some regions. It can be inferred that in practice, government data openness, information openness, and government openness will continue to be confused for a long time. From the perspective of cities that have practiced government data openness earlier, the introduction of specialized regulations and policies has become inevitable. However, due to the lack of higher-level legislation on government/public data openness at the national level, local governments have attempted to formulate policies that combine government data openness with urban development strategies and specific service applications, such as the "Shanghai Public Data and One Stop Network Management Measures (2018)", "Zhejiang Province Digital Economy Promotion Regulations (2020)", "Hangzhou City Brain Empowering Urban Governance Promotion Regulations (2020)", etc.

(2) The deep application of persuasive and learning tools needs to be developed. From the three stages of development, these two types of tools are usually used as auxiliary tools for command based tools and capacity building tools, with limited functions and types. However, for local governments, persuasive and learning tools are less mandatory and more flexible, and are tools with greater autonomy and innovativeness. There is still much room for exploration. Although the expert system and conference competition tools have been increasingly used in cities, and the expert system is gradually integrating into various aspects of the government's open data ecosystem, there is still significant room for innovation in digital culture cultivation and public participation.

Characteristics and laws of the application of policy tools for government data openness in China

Through the previous analysis, it was found that the application of government data openness policy tools in China has certain characteristics and patterns.

（1） Command based tools have conductivity and conversion capabilities

The guidance and work system (authority) tools in command based tools have strong transmission from the central to local levels. For example, in the first and third stages, when the state provides guidance and encourages socialized utilization of information disclosure, government information resource development and utilization, government information resource sharing and management, and government service reform, local governments also use guidance and work system (authority) tools to require the development of management methods and standardized systems that are compatible with disclosure, open utilization, shared management, and public services. Both conductive tools and persuasive encouragement tools are used to encourage individuals, businesses, and other social forces to enter the fields of information resource development and big data application development at the national and local levels. The conductivity of tools is more reflected in the top-down transmission of central spirit and the research and implementation path of local governments, and less in the autonomy of local policies.

When the central government proposes requirements for information resource management, data sharing, information release, and organizational implementation support based on command based tools, local governments usually transform them into capacity building tools such as information support, talent support, technology support, work systems, and assessment and supervision, which are used together with command based tools. Moreover, the response time and focus of each region are different. This is reflected in the guidance and instructions proposed by the central government in the first and second stages, where local governments gradually cooperate to use capacity building tools. Class A governments focus more on public services and external environmental links, while Class B governments concentrate on "open" links.

（2） The use of incentive tools reflects inheritance, convenience, and innovation

The most commonly used incentive tool is the punishment and disciplinary rectification tool, which has been frequently used since the information disclosure period and has continued to be used in the management of information resource openness and sharing, government transparency, and government services. It mainly functions within government departments and is used in conjunction with capacity building work systems (organizations) and assessment and supervision tools. Based on administrative culture and institutional habits, it also reflects to some extent the convenience of "power". In addition, based on the convenience of "money", monetary incentive tools are often used.

In terms of incentive based monetary tools, government procurement and outsourcing, special funds, capital investment, financial and tax incentives, and financing cooperation are commonly used, with a focus on the data "utilization" link, and are used in conjunction with industrial and market strategies. The slight difference between Class A and Class B cities is that, in line with local fiscal capacity and economic foundation, they tend to use direct monetary tools when their strength allows. This is manifested in Class A cities using more government procurement and outsourcing, special funds, and capital investment, while Class B governments are more willing to use fiscal and tax incentives and financing cooperation to promote industrial development.

The innovation of incentive tools is reflected in the use of existing tools and the attempt to use more professional tools. For example, based on the use of punishment and rectification tools, it is proposed to commend and reward individuals and collectives who have made outstanding contributions in promoting data sharing and opening up, and promoting the public value of big data applications. For example, in terms of the use of licensing tools, due to the government's permission for open use of data, there is no operational legal text reference to regulate the responsibilities and rights of various stakeholders. At the same time, the use of licensing tools has not been highlighted in previous institutional designs for information disclosure and information resource development and utilization. Therefore, government data open licensing agreements developed around data request rights, revenue rights, payment (or free), data security and prevention of abuse, data explanation, exemption and exemption have become policy tools for a small number of local governments to try innovation.

（3） Capacity building tools are converters

The guidance and instructions, as well as the work system (authority) in command based tools, are transformed into the work system (organization) and assessment supervision in capacity building tools. This is manifested in the central and higher-level government departments issuing regulations, plans, guidance opinions, and management methods, proposing requirements and guidelines for infrastructure construction and institutional arrangements related to government data openness. Local and lower level government departments simultaneously develop policy documents with lower effectiveness levels but stronger operability, such as work plans and action plans, which are implemented through work systems (organization) and assessment supervision tools. This rule also applies to observations of temporal changes. For example, in the first stage, it was proposed to "strengthen the leadership of the 'top leader' in information sharing and form a long-term mechanism", while in the second stage, systems such as "establishing information sharing supervision and inspection, assessment and notification, security and confidentiality review" were established.

The expansion of the effectiveness of persuasive tools sometimes relies on the cooperation of capacity building tools, that is, the promotion and encouragement of work needs to rely on the coordinated use of information support (information), public services, talent support, technology support and other tools. For example, to encourage the "social utilization" of public data and government information resources, the government often needs to provide places (public libraries, archives) for promoting and publishing such information, publish relevant content introductions, make more people aware of information through official websites, media platforms, etc., and provide certain popular science training to improve users' data and information literacy.

（4） Persuasive tools are both signal lights and probes

Persuasive tools have the function of promoting and encouraging the public. On the one hand, they convey the willingness and attitude of government departments towards data openness and information resource utilization to the public. On the other hand, they obtain feedback information on public needs through online media, public service ratings, and other means. However, persuasive tools have weaker coercive and binding forces, reflecting more the influence of value orientation, and their effectiveness is not easily noticeable in the short term. And there is a significant difference in the effectiveness of using this tool within government departments and among non-specific members of the public. Within government departments, due to the authority formed by the administrative system itself, the target audience for promotion and encouragement is relatively concentrated and clear, and the effect can be expected. However, non-specific members of the public are greatly influenced by factors such as policy interpretation, service support, and public literacy, and the effect is difficult to predict. Therefore, from the perspective of application background, it is usually a tool used by the central government or higher-level departments when launching a certain action strategy or service reform, hoping that the public and social forces can participate together, but without specific supporting measures.

（5） Learning tools are both catalysts and practical tools

Learning tools have the characteristic of diversified participation, and in principle, they act on non-specific objects. However, due to the requirements of data capability and cultural literacy for government data openness and utilization, the target group for this tool to truly play a role is often individuals or enterprises who have a strong interest in government data openness and utilization and have a certain knowledge background. Conference competitions and expert systems are typical examples of such tools, which further enhance their interests and skills through group activities, thereby making them more willing to participate in government data open sharing activities. At the same time, learning tools have a good catalytic effect, such as the "competition driven use" of conference events, which greatly helps improve the professionalism and quality of public services, information support (information), technology support, and industrial strategy tools. For example, the expert system can be well integrated into regulatory standards, work systems (authoritative/organizational), and the use of public service tools.

The use of learning tools can reflect the autonomy and innovation of local governments. Usually, work documents with lower effectiveness levels such as meeting notices and activity notices can be released at the local level, with lower administrative costs, but higher organizational costs and platform integration requirements. While testing the organizational ability and appeal of local governments, there is also significant room for independent innovation. For example, currently, learning tools still have a lot of room for development in stimulating non professional public participation.

In summary, China's government data openness has entered a stage of rapid development, and overall, the types of national and local policy tools are becoming saturated (a total of 32 tools), and the number of application links has also increased. The problem of "dead" information resources and information silos left over from the first and second stages has been resolved; The issues of data management standards and related concepts differentiation have also been addressed in the form of standardized work plans and management measures in the pilot regions; There is always a problem of low social participation and narrow influence, and there is still room for improvement; The application of persuasive and learning tools still needs to be explored; With the gradual enrichment of policy tools at the local level, the differences in the use of policy tools between Class A and Class B governments are becoming increasingly prominent.

In addition, during the process of sorting out policy tools, this study also found that due to the lack of clear distinction between concepts such as government information resources, public data, and government data at the national level, the boundaries of government data openness are relatively broad, and local governments have certain autonomy and interpretation rights. From a practical perspective, local governments not only explore the construction of data open platforms, but also involve data support for government informatization and public services. There are differences in the implementation path and development strategy of local government data openness. In the future, the use of policy tool theory to study the implementation of local government data openness policies still needs to be explored in terms of implementation paths and development strategies.