Abstract:
As the "raw material" and "processing object" of digital economy , data is becoming more and more important in economic development. With the development of information technology, almost everything is being digitized, and data is almost everywhere. As a new factor, data has some fundamental differences compared with land, capital, labor and technology, among which the most prominent features are non-competitive, complementary, external and exponential proliferative. And data element pricing is the core of the data value chain. The data pricing model is very complex and requires full consideration of the balance between efficiency and privacy and security. In addition, realizes open sharing of data elements, which is crucial to promoting sustained economic growth and transformation. However, in the actual utilization of data elements, there are problems such as " data island ", "data chimney", "data monopoly", and "data black market". Only by achieving a balance between efficiency, fairness and privacy, meeting feasible property rights recognition, effective privacy protection, reasonable income distribution mechanism, necessary key information and other prerequisites, and effectively combining government rules, social and market spontaneous forces, can we promote the effective allocation of data elements.
1
What is the data
As a new element, data is increasingly valued. So, what exactly is the data? What are the main characteristics? There are many analyses in both the academic and practical circles. An analysis framework that can be learned from is Ackoff (1989) proposed the DIKW framework, where D represents data (data), I represents information (Information), K represents knowledge (Knowledge), and W is wisdom (wisdom). Afterwards, many scholars (Bellinger et al., 2004; Rowley, 2007; Xu Zhong, Zou Chuanwei, 2020) have expanded and improved this framework and classified the characteristics of data, information, knowledge and wisdom.
data is a symbol (Symbol), record, and the specific form can be an article, a sound, or a picture. International Data Management Association believes that "data presents facts in the form of text, numbers, graphics, images, sounds and videos" (DAMA, 2020). Data is often generated from the interaction between people, people and things, and people and nature. In the era of the Internet of Things, the Internet of Things is popular, and interaction between things will also generate massive data.
and the information is processed useful data for eliminating uncertainty. Information refers to facts or details that can answer questions such as what environment, who is, what happened, etc. For example, tomorrow will be very hot. This passage is a piece of information that can transform the uncertainty of tomorrow's weather changes into a certain certainty, which is also the value of information. In " Understanding the Process of Economic Change " (2007), North also distinguishes information and knowledge when discussing uncertainty. Given the existing knowledge stock, uncertainty can be reduced by increasing information.
Knowledge is the accumulation of material and social environmental laws and patterns. It is an organized or logical explanation based on data and information, and can create new value. Knowledge can be divided into coding knowledge (explicit knowledge) and uncoding knowledge ( implicit knowledge ). From information to knowledge, it is a sublimation of human cognition and a process of turning from external experience to introspective perception. There is no doubt that although data or information is massive, it requires more effort and investment to transform into perception of the world and knowledge to transform the world. Obtaining algorithms and models through massive data or information can be regarded as a kind of knowledge. In other words, knowledge is similar to production functions, and data is similar to an input element of production functions (Jones and Tonetti, 2020)
Following the analysis paradigm of Romer (1990), Aghion and Howitt (1992), this article further simplifies the above-mentioned DIKW four classifications into two dimensions: data and knowledge. Among them, data is a factor input, while knowledge is a production function. The subsequent discussion will generally follow this framework.
2
Basic characteristics of data elements
(I) Non-competitive
data is easy to store and has strong replicability. There is almost no limit on replication if the operation, maintenance and storage costs are not considered. Whether it is the survey data released by statistical agencies or the personal data collected by various Internet platforms, it can be used by multiple users at the same time. In a physical sense, there will be no losses, and there will be no mutual influence. The non-competitive characteristics are obvious. A typical example is the machine learning competition established by Kaggle, where relevant data can be used by different contestants. Therefore, data cannot be simply compared with natural resources such as oil, because the latter is competitive. If one more barrel of oil is exploited, the earth's oil storage will be reduced by one barrel (Varian, 2019). Instead, it seems more appropriate to compare it to sunlight. Of course, during the actual use process, the acquisition of data requires necessary efforts and conditions, and is not as cost-free as obtaining sunlight, so the data still has certain exclusive characteristics (Partially Excludeable; Carriere-Swallow and Haksar, 2019). Because of this, according to non-competitiveness and certain exclusiveness , data can be regarded as a quasi-public product or public product.
(II) Complementarity
Data from different sources can be integrated with each other, which can improve the ability to reveal hidden clues or rules, increase the marginal value of a single data source, and to a certain extent, digital elements can have the characteristics of scale rewards that remain unchanged and even increase. Remuneration of scale remains unchanged and network externalities are important sources of economic endogenous growth in . Krugman and Romer are one of the few mainstream economists who recognize increasing returns, and Arthur of the Santa Fe Institute has systematic research on this (Complex Economics, 2018).
There is also a view that as an input element, data still has a decreasing feature of marginal returns, which is no essential difference from other elements. Varian (2019) gave an example to illustrate that the accuracy of image recognition will increase as the amount of data fed during training increases, but the speed of improvement will gradually slow down. But such a view only emphasizes the single application of data elements. It is particularly worth noting that due to the development of digital technology , behaviors or things that were originally difficult to digitize can be digitized, truly realizing the "everything is a number" in ancient philosophy, which creates unprecedented conditions for cross-border digital fusion and enables the originally irrelevant fields to be connected. This is also an important feature that different from the industrial age in the digital age. Data is non-competitive and can be used for multiple purposes, and the marginal cost of subsequent use will continue to decline. Coupled with complementarity and networking and learning effects, data elements may still have the characteristics of increasing returns to scale.
(III) Externality
Data sharing does help promote R&D, improve product and service quality, and improve efficiency. However, at the same time, due to the existence of information asymmetry and the market monopoly of the platform, the party sharing the data may not be able to obtain sufficient compensation, which creates privacy externality (Carriere-Swallow, Haksar, 2019). For example, consumer data may be transferred to third parties without knowing it, resulting in more spam harassment or unfavorable price discrimination against (Odlyzko, 2003). These negative externalities are difficult to internalize by consumers and data acquisition companies.
Since different data are related, linear planning can be used to infer unknown data from a type of known data, and can infer the behavior of another group of users from behavior data . This may cause user privacy leakage during the data disclosure process, causing data ethics issues. It will also lead to oversupply of data and low data prices, thereby reducing the efficiency of the data market (Acemoglu et al., 2019). This also determines the secure shared use of data elements, which is closely related to the development of encryption technology.
In addition, digital enterprise may rely on network externalities, over-concentrate data, and obtain a large amount of excess returns.Mastering or owning unique data is just like mastering unique patents and technologies, which can enable the controller of data to obtain an excess profit. In this way, the private attributes of the data will be strengthened and the sharing and integration of data will be hindered. To a certain extent, applying for patents and obtaining unique data are important ways for market entities to establish competitiveness or moats.
(IV) Exponential proliferative
data can be regarded as an accessory to various conscious or unconscious activities. The application process of data itself can generate more new data and speed up decision-making or algorithm iteration. For example, driverless car is controlled by an algorithm based on data training. The longer the mileage, more scene data will be generated, which in turn can promote further optimization of the algorithm and form a self-accumulation and growth process of "data-algorithm-data". With the improvement of the degree of digitalization of the entire economy and society, the popularity of smartphones, the wide application of sensors, the upgrade of broadband transmission technology, the continuous enhancement of computing power, global interconnection, Internet of things, and data elements are showing an exponential growth trend. According to IDCh estimates, global data is expected to reach 175ZB by 2025, an average annual growth of 27% compared with 2019, which is equivalent to double the data scale in three years. Data scale is a function of increasing economic scale, and super-large economies may accumulate more data advantages as a result. There is some controversy as to what impact the exponential growth of
data will have on economic growth. One view is that data increase has a growth effect, that is, the increase or unchanging return on scale of data can achieve sustained economic growth. Due to the existence of a data feedback loop (Data Feedback Loop), enterprises can obtain greater market position and thus obtain more data (Farboodi et al., 2019). The huge threshold effect and network externalities help achieve economies of scale. In addition, the generation of more types of products and services around digitalization has expanded the existing product and service space, thereby promoting sustainable economic growth. But there are also objections, believing that the increase in data will only have a horizontal effect (Level Effect t), or that is, due to the decreasing marginal returns, more data will not change the path of economic growth. Even if there is an increase in return on scale, this effect only exists within the enterprise and cannot be applied to the entire economy. Bajari et al. (2018) used Amazon data to conduct empirical evidence, indicating that there is an upper limit on the benefits from increasing data size.
3
Data elements How to price
Data As an element, its value realization key lies in connection fusion and open sharing. Combining data elements and other elements will inevitably create new value. The data value chain involves the generation, collection, transaction, and use of data, among which the pricing of data elements is the core of the data supply chain. Before discussing data pricing, you need to distinguish between data products (Data products) and digital product (Digital products). The former refers to the objective historical records of human activities, such as census data and commercial databases. The latter is a product or service presented in digital form, such as an e-book, a digital movie, a digital map, a digital production solution, and so on. More new data will be derived during the production and use of digital products. Below, the realization of data value will be discussed from the perspectives of data products and digital products.
Figure 1 Data supply chain diagram
(I) Data product
From the perspective of the data supply chain, data sales are basically at the front end of the data supply chain. Before the specific data is sold, it may involve data collection, cleaning, storage, visualization and other links. Data companies like Bloomberg, Wind, CEIC, Dun and Bradstreets are mainly engaged in collecting, organizing, and collecting data from different sources or specific fields, and then selling them to downstream customers.Some of the relevant data come from traditional statistical or survey institutions, such as government statistical institutions and industry associations, some come from the production and operation conditions of enterprises, and some are surveys of various personal behaviors. In terms of scale, more data comes from e-commerce platforms and industrial Internet platforms. Data customers include financial institutions, government units, enterprises, scientific research institutions, etc. In reality, there are also some informal or even illegal data transactions, which are often difficult to obtain through formal and public channels. The value of
data is often related to the characteristics of the data substitutability, update frequency, data granularity, integrity, availability, etc. But unlike general commodities, the marginal cost of data replication is close to zero, and data pricing depends more on the value evaluation of the demand side rather than the cost bonus, and the data pricing model is more complex. Sometimes, the value created by data users or the scope of value realization, in turn determines data pricing, which is manifested as post-event pricing. At present, common pricing models include free, free + paid value-added, on-demand charges, fixed rates and other methods. Among them, most public data from government institutions or other public institutions are free of charge. The free + paid method is mostly to use free data to attract potential customer groups and pay to purchase higher value data or data value-added services. Free data is usually not timely and has a coarse particle size. The on-demand charging method is generally to transmit data through the API interface and charge according to the specific usage, which is commonly found in high-frequency scenarios such as finance. The fixed rate is when the customer purchases an account and then obtains the permission to use the data within a certain period of time. The latter two methods can also be combined to form two pricing models, which can achieve greater revenue maximization and are also more common in general data sales (Wu and Banker, 2010).
With the development of the digital economy, it is more convenient to collect and utilize information from micro individuals, but at the same time, people's expectations for strengthening individual privacy protection are also more urgent. Data pricing needs to fully consider the balance between efficiency and privacy and security. As mentioned earlier, data is not an ordinary private item. If non-competitiveness and privacy externalities are not considered, data pricing will be distorted. For example, if a query to be performed (Query) is a linear combination of other queries, or is inferred through other queries, there is arbitrage behavior to purchase this query. For example, due to the existence of social networks, the personal information of different consumers is often related. By purchasing some consumer data, you may be able to infer other consumer information. Because of this negative externality, data oversupply will occur, resulting in the data value being seriously underestimated (Accemoglu, 2020).
Figure 2 Arbitrage-free pricing for data elements
A area that is currently discussed is Arbitrage-free pricing. Figure 2 shows a schematic diagram of arbitrage-free pricing. The equilibrium value E of data elements depends on the balance between privacy and data value. On the one hand, the willingness of the data purchaser to pay changes with the quality of the data, that is, the value curve. The data quality is higher, and the information can be revealed is more valuable. On the other hand, data owners (such as individuals) generally value privacy. If they disclose personal information, they need to receive corresponding compensation, that is, the compensation curve. To provide more private real information in terms of data ownership, higher compensation is required. This pricing method goes beyond the original approach of absolute privacy protection and effectively combines privacy protection with data use. At the equilibrium point E, the price can not only provide appropriate compensation for data from different sources based on the data quality, but also overcome the arbitrage behavior in the data purchase process. Of course, to achieve this arbitrage-free behavior, it is also necessary to develop related encryption technologies, such as privacy computing.
As a privacy calculation method, secure multi-party computing (MPC) has attracted much attention in recent years. With the popularity of machine learning and artificial intelligence , the training and application of an algorithm often requires the use of data from different sources, which involves the issue of how to protect their respective privacy and security while determining the contribution of data from different sources.Secure multi-party computing can break the data silos, realize controllable sharing of data, and minimize the risk of data leakage, which has important theoretical and practical significance. Security multi-party calculation was proposed by Yao Qizhi in 1986 (Yao, 1986). After entering the Internet and digital era, the demand for data sharing is more urgent, and secure multi-party computing has also been further developed. By 2018, major technology companies such as Google , Alibaba have achieved some business cases (Hong et al., 2020).
Of course, many companies that master data or big data do not simply sell data, but more by providing data-related value-added services, and the data itself has not been transferred. E-commerce platforms, social platforms or basic telecom operators have a large amount of user information and can sometimes use data to accurately portray individuals, enterprises or institutions, thereby providing data purchasers with services such as traffic diversion and price discrimination. For example, an advertiser can target advertising through bidding rankings on search engines and pay a certain fee. For example, Internet platform companies can promote information sharing through the industrial chain, and Cainiao Logistics connects data with different express logistics companies to optimize logistics paths to solve the problem of logistics congestion during the "Double Eleven". By using big data to accurately locate the targets of policy roles, the government can also further improve governance, such as identifying low-income groups to issue consumption coupons, identifying active market entities to implement targeted support policies, etc. Many companies provide data analysis services by integrating data and integrating their own intelligence through artificial intelligence and machine learning methods. For example, some big data companies provide big data credit reporting services. This type of service pricing involves the data value-added part, and the pricing method is more difficult to unify.
(II) Digital product
From an economic perspective, digital products have the characteristics of winner-takes all, high fixed cost-low variable costs, more emphasis on experience, and more diversified income monetization channels.
4
Who should own data: From property rights to accessibility
Usually, clear and clear definition of property rights is the basis for the effective allocation of resources. Coase believes that initial property rights allocation is not important. As long as there is sufficient competition and the related benefits and costs can be fully internalized, resources can be effectively allocated (Coase, 1960). If privacy is also regarded as a right, it does not matter how privacy rights are configured between consumers and companies that serve as data collectors. This is also the view of Chicago School based on property rights (Laudon, 1997). However, from the characteristics of data elements that are non-competitive, private external, complementary, etc., excessive data collection, infringement of privacy, and use data advantages to seek market monopoly power frequently. To a certain extent, Coase Theorem is not applicable to data elements. Of course, if data use is strictly restricted due to privacy protection, the economies of scale of data cannot be used (Jones and Tonetti, 2020). In practice, balancing multiple goals such as efficiency, fairness and privacy to achieve effective use of data, the ownership of data ownership is not black or white, but it is more likely that there is a broad spectrum between the two views mentioned above.
data property rights distribution has a significant impact on welfare. Data is a by-product of economic activities, and if a company has the personal data obtained from the process of dealing with customers, it will help it increase its investment in data collection, analysis and utilization. The downside is that companies do not necessarily respect customer privacy. At the same time, due to concerns about creative disruption, he is even more reluctant to share data with other companies. With data advantages, companies can have greater market power and may hinder other companies from entering, affecting competitive fairness. Studies have shown that if the non-competitive characteristics of data cannot be exerted and the economic effect of is realized through connection, fusion and sharing, , the loss of welfare costs in the whole society will be huge. In particular, the complete ban on data sharing will lead to a reduction of social welfare by nearly 60% compared to the optimal level (Jones and Tonetti, 2020).From the perspective of current theoretical research and practical development, if data elements belong to consumers, consumers can weigh the privacy benefits and the economic benefits brought by exchanging personal data on their own, thereby achieving more effective allocation of data elements and improving the welfare of the whole society. In recent years, financial regulators have begun to realize this, granting individuals more data rights, actively promoting the development of Open Banking, increasing the open sharing of financial data, and promoting financial innovation and competition.
Private items emphasize ownership ownership, but data elements are non-competitive and partially exclusive, have stronger public or quasi-public goods properties, and access (Access) may be more important (Varian, 2018). Excessive emphasis on the ownership of data elements will limit the flow, sharing and reuse of data and cannot release potential value. At the same time, some designs are not necessarily practical, such as "portability of personal data". The data subject not only has the right to know, access, and correct the data collected by the data controller, but also has the right to transfer this data to a third party. But this may lead to a conflict between property rights and personality rights. Unlike the data's personal rights and interests ownership, there are many controversies in various aspects. To this end, data accessibility is an important variable in promoting market competition and fully leveraging the value of data in the digital economy era (Cremer et al., 2019). In fact, the EU , which strictly protects personal privacy, realizes that the development of the digital economy lags behind the United States. After the implementation of GDPR, it has further formulated the EU Data Strategy and launched the Digital Services Act and the Digital Market Act, trying to promote the open sharing of personal and non-personal data while ensuring data security.
enhances data portability and the conversion convenience of different data infrastructure interfaces, helping to improve accessibility. In particular, the Internet of Things is accelerating its development, and it is becoming more urgent to achieve safe and feasible data exchange channels based on interfaces that are universal and highly standardized. In specific practice, there are already application models such as data space, data banking, and My data. However, it is also necessary to note that because data has non-competitive characteristics, strengthening data sharing may further strengthen the data advantages of incumbent companies due to scale effects and network externalities, which in turn will damage competition.
5
How to promote open sharing of data
Data elements are a basic strategic resource, realizing open sharing of data is crucial to promoting sustained economic growth and transformation. However, in the actual use of data elements, there are both "data silos" and "data chimneys", which are blocked by departmental interests and are seriously wasted public data resources; there is also "data monopoly", in which incumbent companies abuse their data advantages and block the data value chain; there is also a "data black market", personal information is over-collected and privacy is not effectively protected, resulting in serious social ethical problems. Only by achieving a balance between efficiency, fairness and privacy, meeting feasible property rights recognition, effective privacy protection, reasonable income distribution mechanism, necessary key information means and other prerequisites, and effectively combining government regulations and spontaneous social and market forces, can we promote the more effective allocation of data elements.
(I) Give full play to the first leverage effect of government big data, accelerate the pace of data openness and sharing
Government functional departments and some public institutions have accumulated a large amount of data in their daily affairs, involving all aspects of economic and social operation. The data source is stable, the scale is considerable, and it has good authenticity, integrity and complementarity. With the development of e-government, the government has mastered about 80% of the information resources of the whole society. Government big data has the strongest public attributes, and there are relatively few disputes and disputes about who should have data. The government’s open data has strong demonstration significance and can better promote the open sharing of data in the whole society. The United States launched the DataGov website in 2009 to increase government data opening. In 2019, the Federal Data Strategy and 2020 Action Plan was released to promote data protection, sharing and opening up.The UK also mentioned in its digital strategy that it would change its use of government data. Domestic digital government affairs are also promoting rapidly, with some good cases among them. For example, Zhejiang established a "One-time at most" reform office to connect data from public security, social security, real estate, taxation, education and other departments, so that the data can run more errands and the people can run less errands.
But on the other hand, due to the imperfect data sharing and opening system of functional departments and incomplete information infrastructure, there is still a lot of room for further opening up the data link. Administrative barriers between different government functional departments are often difficult to break, data is difficult to be shared, or it is limited to statistically-level, relatively outdated limited data sharing. When obtaining data across departments, sometimes after coordination by the main responsible person, various functional departments can still refuse, delay or reduce their provision. In addition, the data from many places are collected together, and due to limitations such as assembly and funding, development and utilization are far from enough.
Next step, we should take digital government construction as the top leader project, and improve the communication and coordination mechanism for data sharing between different levels, different departments, and functions. Issuing and improving the data sharing list, expanding data opening in steps and levels, and establishing relevant performance appraisal mechanisms. Strengthen digital transformation training, promote good practices in various places, enhance data sharing awareness, and enhance the digital skills of government departments. Strengthen the standardization of government data and information collection, processing and storage, improve the charging mechanism for the use of government big data, improve the data sharing infrastructure, and enhance security protection capabilities.
(II) Cultivating the third-party data market and strengthening the data industry
Cultivating and strengthening the third-party data market is an inevitable requirement for deepening the division of labor in the digital industry chain and fully releasing the value of data. During the actual use of data, from data generation and collection to the final demand side of data, there may be multiple links in between, and each link needs to have corresponding professional knowledge. For example, between the bank and the lender, the lender can directly provide information to the bank, and the bank can decide whether to issue a loan based on the lender's information. But at the same time, banks can also purchase loan credit information from third-party credit reporting companies to assist in decision-making. Credit reporting companies undertake the functions of collecting, sorting and analyzing data. Of course, the data mining tools or technologies used in the relevant links come from other companies.
Domestic third-party data has developed rapidly in recent years, and a group of relatively professional big data companies have emerged. Some places have also established big data exchanges or trading centers, and have made certain progress in data development and utilization. But overall, the development of domestic data service companies is relatively lagging behind, with more than 3,000 but small in scale. There are also lack of data leaders such as Reuters, , Bloomberg, , Dun-Brathestre, and RelX. Data companies like Bloomberg and Lixun all have annual revenues of more than US$10 billion. The construction of the data market is not perfect, the market activity and participation are not high, and the value of data is not fully released. A large amount of data illegally enters the underground data transaction chain, and data security incidents occur frequently, and the privacy rights and interests of enterprises and individuals are not effectively protected.
Next step, we must further encourage enterprises to improve their comprehensive management capabilities of data elements, and build a data sharing platform through industry organizations, formulate data sharing standards, and continuously increase the supply of high-quality data elements. Fully consider the characteristics of data elements, actively explore data sharing methods such as trusted third parties, Paid-in-Kind, club model, and data market to promote the deep integration of data elements with various specific scenarios. Accelerate the research and development of computing technologies for privacy protection, promote the combination of privacy computing and blockchain, and realize that data is "available and invisible" during the open sharing process to meet more complex and diverse data needs. Through taxation, government procurement, finance and data opening and other measures, we actively cultivate data service enterprises.
(III) Strengthen domestic data legislation and supervision, improve data governance level
As data security governance has become a focus that cannot be ignored, major economies have strengthened data legislation practice and strengthened the protection of information rights. The EU's General Data Protection Regulation (GDPR) provides an example for protecting consumers' data rights. In May 2018, the EU GDPR officially came into effect, aiming to provide legal guidance for various types of enterprises and institutions to collect and utilize personal information in their business activities. Compared with previous data protection regulations, GDPR expands the rights of data subjects (Subjects) and increases the obligations and responsibilities of data control and management parties. GDPR gives data subjects seven rights, the most eye-catching of which is the right to deletion or the right to forget (The right to erasure). When one of six situations such as "personal data is no longer necessary to achieve the relevant purpose of its collection or processing", the data subject has the right to require the data controller to delete their personal related data in a timely manner. Due to the large economic size of the EU, GDPR also produced a significant spillover effect .
EU legislation has a certain demonstration effect. California, India, and , Brazil, have also gradually begun data legislation after the GDPR came into effect. my country is also actively improving legislation on personal information and data power from the national and regional levels. Among them, on October 21, 2020, the Legislative Affairs Committee of the Standing Committee of the National People's Congress issued a draft for soliciting opinions on the "Personal Information Protection Law of the People's Republic of China (Draft)" and solicited opinions from the public on legislative issues related to personal information protection. The practice of data legislation in different economies reflects to a certain extent the individual's rights and claims on data more prominently. The local level has also strengthened the specifications for data utilization. On June 29, 2021, Shenzhen passed the "Shenzhen Special Economic Zone Data Regulations", covering personal data, public data, data element market, data security and other aspects. It is the first basic and comprehensive legislation in the domestic data field, and has proposed data rights for the first time, clarifying that individuals enjoy personal rights and interests in data, and that enterprises enjoy property rights and interests in products and services formed based on data.
Next step, data legislation and supervision also need to handle the triple balance. First, the balance between development and security. my country's digital development is at the forefront, and there are many challenges it has encountered. In terms of data specifications, we should put forward more creative proposals based on domestic actual conditions. The digital economy is a must-fight place for major economies in the world. The value of data cannot be fully utilized, the development of the data industry is lagging behind, the digital competitive advantages are insufficient, and the digital momentum is weakened, which is the biggest risk. The second is the balance between technological innovation and technological ethics. The digital field is one of the most active areas of innovation at present, and relatively inclusive and prudent supervision is needed to increase fault tolerance and encourage more exploration of digital frontiers. However, while the applications of technology such as big data, algorithm recommendation, face recognition bring convenience to everyone's life, they also bring challenges to privacy protection. Personal privacy protection requires the implementation of data application and governance. The third is the balance between government and social forces. Data utilization often involves multiple subjects, multiple links, and multiple sources. While filling in the shortcomings in government supervision and clarifying data ownership and competition rules, it is necessary to give full play to the role of social organizations such as industry associations and alliances, strengthen industry self-discipline and standardization, improve enterprise data management capabilities, and reduce security risks.
(IV) Build a high-level cross-border data flow policy system to avoid being surrounded by "rules"
The value of data is in flow, connection, integration and sharing. Promoting the orderly and convenient cross-border flow of data elements is an important prerequisite for maintaining the innovation chain of the global digital supply chain industry chain, improving the level of investment, trade and business environment, and promoting global digital cooperation. Studies have shown that bilateral digital ties have increased by 10%, and trade in goods and services will increase by about 2% respectively. If combined with the Regional Trade Agreement (RTA), trade volume can also increase by an additional 2.3% (Lopez-Gonzalez & Ferencz, 2018).But at the same time, outdated regulatory rules, data flow restrictions, local existing requirements, network security risks, intellectual property protection, etc. hinder digital connections and curb digital innovation.
Promote the orderly and convenient flow of data elements, which is inseparable from international cooperation. However, there is no global cross-border data rule system, and the existing cross-border data regulation has shown the characteristics of clubbing. If we look at the dimensions of privacy protection, corporate competition, digital innovation, national security, etc., the cross-border data flow regulations of various countries are generally divided into four categories. The first category is that the United States promotes cross-border privacy protection rules (CBPR) under the APEC framework, which emphasizes the free flow of data and globalization. The second category is the EU's regulatory system based on GDPR and subsequent bills, which emphasizes more on personal privacy protection and data localization. The third category is some developed and emerging market countries, which emphasize localization more and at the same time try to get closer to the regulatory standards of the EU or the US. my country can be classified into a separate category, generally emphasizing data localization, but it also advocates data security and orderly cross-border flow, and adopts a localization + security assessment mechanism.
From a trend perspective, although Europe and the United States have major differences on cross-border data flow rules, we must also see that coordination between Europe and the United States is constantly strengthening. In February 2021, G7 issued a joint statement, which will promote the free and trustworthy flow of data and improve the governance of the digital economy. In addition, Japan, Singapore, Switzerland and other countries have gradually moved closer to Europe and the United States in terms of rules, hoping to further integrate into the free flow circle of data. OECD's research shows that from a country comparison, my country's digital trade restrictions are the highest among the 44 sample economies, and faces the risk of being surrounded by the "wall of rules" and being limited in the competitiveness of the digital economy. It is necessary to combine internally and externally to actively build high-level cross-border data flow rules. It is necessary to improve the classification and hierarchical management system for cross-border data flow and further clarify specific operating guidelines such as data security assessment standards and procedures. Give full play to the institutional convenience of the free trade zone port and carry out pilot projects on cross-border data flow rules. Make full use of multilateral platforms such as G20 and WTO, and take advantage of the implementation of regional trade agreements such as RCEP and the possibility of joining CPTPP negotiations to participate in the formulation of international data flow rules to better balance data flow and data security.
Source: The book "Digital Macro: Macroeconomic Management Changes in the Digital Era" Chapter 6 (Authors: Chen Changsheng, Xu Wei)
Copyright belongs to the original author. The sharing is for learning reference only. If there is any inappropriateness, please contact us to deal with it.
