Huang Yiping: The platform's economic data is safe and cannot be separated from the coordination of governance policies

On July 21, 2022, the National Internet Information Office was fined RMB 8.026 billion in accordance with laws and regulations such as the "Cyber ​​Security Law", "Data Security Law", "Personal Information Protection Law" and "Administrative Penalty Law" and other laws and regulations. The chairman and CEO of Didi Global Co., Ltd. Cheng Wei and President Liu Qing were fined RMB 1 million everywhere, causing public attention. The "Didi Incident" is not simply a country -the incident of capital relations. It reflects the profound tension between capital elements and national capabilities in the era of new technology revolution. The "Didi Incident" is characterized by data security, but behind it is the parallel and mix of data security and capital monopoly logic.

At present, data as a new type of production is deeply affecting the development of the country's economy and society. Data security guarantee capabilities are a direct manifestation of national competitiveness. Data security is an important aspect of national security, and it is also an important issue for promoting the healthy development of the digital economy and enhancing national governance capabilities.

In recent years, it has become a very common phenomenon such as big data abuse, user information leakage, data theft and trafficking. How to maintain data security in the economic development of the platform and better give play to the positive role of platform economy in promoting economic growth and improving people's lives, which is a subject worthy of in -depth research on the government and platforms. In the "Platform Economy: Innovation, Governance and Prosperity", Professor Huang Yiping discussed the importance of data security to the platform economy and made several suggestions for the security governance of the platform's economic data:

---

Data is new coal

The importance of data can be confirmed from the popular saying "data is new oil". However, Neil Lawrence, who had served as the head of the Amazon machine learning team, believes that "data is new coal" is more appropriate, because the steam engine during the first industrial revolution is more useful for those who have a large amount of coal. Without data, there will be no platform economy. For the first time, the Fourth Plenary Session of the 19th CPC Central Committee proposed to the distribution of data as a production factor. This is a major theoretical innovation, marking the independent production of data from technical elements to become a separate production factors.

The efficacy of data analysis has been fully demonstrated in the field of platform economy: the car -on platform can match taxis and passengers at the lowest cost and shortest time; the takeaway platform can ensure that the takeaway is sent from the restaurant to the restaurant at the fastest speed to Ordering people's hands; short video platforms can be pushed to related mobile terminals that reflect personal preferences at any time; digital financial platforms can analyze users' online shopping, social and other digital footprints to provide credit services and control credit risks. In addition, data analysis can also help optimize urban traffic management, identify new crown viruses, and even track criminals. In short, data analysis can help platform enterprises to improve operating efficiency, change business models, and enhance personalized services.

But at present, there is still a long distance from the data to the production factors. First, you need to clarify the relationship between the data, including ownership and right to use. However, unlike traditional production factors, it is very difficult to define clear ownership and use rights for data. The data itself has private, platform and public triple attributes. At present, the application of various platforms is mainly "unique" data, not big data in the strict sense. The key to forming big data is to break the island and achieve sharing. Compared with traditional elements, the advantages of data elements in configuration are that they can be reused, and the disadvantage is that it can be reused. Anyone can use a set of data without scarcity, but it also involves a complex interest game, which makes difficulties in transactions and pricing. In turn, it will affect the collection and analysis of data.

---

The three major problems existing in data analysis in my country

In the past, my country has not formed a complete set of data governance frameworks. The freedom of the collection and data of platform companies is relatively large. The advantage is that various innovative activities that use big data analysis are very active. universal. In summary, there are three major problems in data analysis:

First, personal privacy and rights have not been protected.

Platform enterprises and other platform service providers have picked up data from other websites, collect personal information of mobile phone users, and even open mobile microphones and eavesdropped talks without permission. These are illegal acts. In August 2021, the National People's Congress passed the "Personal Information Protection Law of the People's Republic of China". It is believed that my country will change huge personal data protection. But the current greater challenge is how to protect the necessary personal rights while protecting the accumulation and analysis of data. Of course, the disorderly phenomenon of the past needs to be changed, but if it takes very strict protection measures to cause the digital economy (including the platform economy) to shrink, it is not an ideal result.

Second, the configuration of data as a production factor still needs complete policy framework support.

At present, the so -called big data is basically the data of several large platform companies, but even such a limited "big data" has played a very large role. So is the data scattered in different platform companies, or is it concentrated? If the feasibility of the data set is not high, the issues of effective sharing must be considered. In fact, some static data, such as taxation, rent, water and electricity bills, as long as I have authorized, sharing is relatively easy, and it is simple to use. And some dynamic data such as shopping, social, and search data are originally non -standard information. I do n’t necessarily understand the content of the information, and it is difficult to share and use it. In fact, there are already mature technology that can support the sharing and use of data on the premise of protecting privacy, such as federal algorithms. But the greater challenge is the policy level: such as confirmation, who the data belongs; such as pricing, how to evaluate the value of data; and so on. Third, the use of algorithms also requires a set of rules that can be accepted by all parties.

It should be said that big data algorithms are actually an important source of platform economic productivity. With a good algorithm, it is possible to accurately market, control risks, and improve user experience. Large technology credit in digital finance is to use big data and algorithms for credit risk assessment to issue loans for enterprises and individuals who lack financial data and mortgage assets. However, there are also many problems with data algorithms. What has suffered the biggest criticism is the algorithm black box (opaque algorithm goals, intent, and responsibilities) and algorithm discrimination (using the algorithm to automatically achieve differences). A objective problem is that although big data can help the platform reduce the degree of information asymmetry, thereby better optimize the process, match transactions, and control risks, but the platform participants, including e -commerce, taxi drivers, outsses and consumers For speaking, the degree of information asymmetry may increase. In this way, the platform's pricing strategy of thousands of people and thousands of faces can easily cause consumers' guessing of discriminatory pricing. This is most likely because of the algorithm black box, or it may be because of the algorithm discrimination or both. Of.

In recent years, relevant government departments are stepping up the improvement of data governance policies. This is a correction of almost self -current phenomenon in the past, but the efforts to build a data governance policy framework may have just begun. The biggest challenge is how to establish a pragmatic framework, which can protect rights and interests, but also exert efficiency. It is not a good choice to take extremes. If you do not protect data, it is easy to cause confusion; excessive protection may kill innovation activities.

The platform's economic data is safe and cannot be separated from policy coordination

---

The platform economic governance contains two levels of content: one is platform governance; the other is public governance, including government governance. The former focuses on how the platform handles the relationship between the platform and the user and the user, while the latter involves the role of the platform in social governance.

As an enterprise, the platform must accept the government's regulation, but the platform must also play a certain regulation function. The goal of platform governance should be fair, fair, and transparent to achieve the orderly development of the platform economy.

1. Improve the legal system in the field of digital economy, formulate digital economic law on the agenda, and provide a systematic legal basis for platform economic governance.

my country has promulgated many laws and regulations related to the governance of the digital economy, including the "Consumer Rights Protection Law", "E -Commerce Law", "Anti -Unfair Competition Law", "Anti -Monopoly Law", "Network Security Law", "Data Security Law", "Data Security Law" And "Personal Information Protection Law" and so on. It is recommended that the National People's Congress promotes the formulation of a digital economic law that can cover all the digital economy fields, improve the connection between existing laws, and clarify the necessary legal boundaries. In particular, it is necessary to explain the important concepts of some digital economy fields. The meaning of the disorderly expansion of capital, avoiding some administrative departments to expand the expansion of the economic field according to their wishes. In the future, the digital economic law can be used as the basic law in the field of digital economy and leads the platform's economic governance practice.

2. Improve the coordination of platform economic governance policies, try to clarify the division of labor and enhance policy coordination in the short term, and consider the establishment of comprehensive platform economic governance agencies for a long time.

my country's platform economic supervision has both industry supervision departments, such as the Ministry of Transport, the People's Bank of China, and the Ministry of Industry and Information Technology, as well as general regulatory authorities, such as the General Administration of Market Supervision. They both manage market order and are responsible for antitrust law enforcement, and also formulate data governance rules. However, the division of labor between institutions is not very clear, and lack of effective coordination, which is easy to cause supervision competition. First of all, you can consider setting up a coordination mechanism at the level of the State Council. It is necessary to eliminate the gap of regulatory supervision, but also prevent repeated governance, and at the same time control the rhythm of the new policy, and try to strive for a smooth transition. The scope of the platform economy means that the traditional industry supervision system is not necessarily effective. In the long run, we should consider establishing a comprehensive platform economic governance agency, so as to comprehensively evaluate the behavior and effects of comprehensive platforms.

3. It is recommended to set up a data governance committee to formulate and coordinate data policies in innovatively. At the same time, the algorithm audit is promoted. Do not simply apply the governance ideas of traditional elements.

Based on the non -competitiveness of data as a quasi -public product and some exclusive characteristics, the traditional methods of confirmation and then trading are no longer applicable. It is recommended to set up a high -specification, cross -departmental, government -enterprise combination of data governance committees to coordinate data policies. Including: the trading scope, algorithm management, personal information protection and data security of data production factors; the application, review, distribution, restriction, and revocation of data licenses; promote algorithm audit; coordinate personal information protection and data security and data security The work; set the dispute solution and coordination mechanism; and so on. It is recommended that algorithm audits focus on requiring relevant enterprises to report input and output and evaluate results. Report input and output refers to the required platform to clearly report the data sources and quality used in different stakeholders' relevant parties, algorithm assessment, and algorithm selection, algorithm prediction or optimization goals, technologies used by algorithms, algorithm operation effects, and so on. Results can include assessments to whether discrimination, validity, transparency, safety, and acquisition. The goal of algorithm governance is to strengthen the interpretability of the algorithm, but also promote the platform to publish relevant rules to reduce information asymmetry.

From an international perspective, the supervision of digital platforms in my country needs to promote innovation and maintain international competitiveness on the basis of protecting the rights and interests of consumers and other data subjects. Regardless of the control of data production factor, algorithm governance or privacy, and data security protection, the regulatory authorities can consider strengthening the transparency of information on both ends of the platform and reducing information asymmetry as the main supervision principle to prevent platform companies from making new information for manufacturing new information. Symmetry, promote enterprises to work hard on increasing bilateral and even multilateral information transparency.

"Platform Economy"

Editor Huang Yiping

---

CITIC Publishing Group in July 2022

Introduction: The platform economy has become a realistic existence that is closely related to everyone's daily life, and has also greatly promoted economic development. However, the disorderly expansion of the platform economy has also brought many challenges.

The platform economy should not be unreasonable prosperity, but to develop in an orderly and standardized manner so that it can benefit the economic form of every ordinary person. Therefore, the governance, innovation and supervision of the platform economy is now imminent and widespread concern.

This book is edited by Professor Huang Yiping of the National Development Research Institute of Peking University, and organizes a number of scholars from many fields such as economy, digital finance, Internet, management, law, and other fields. The impact of platforms on employment and income distribution, data governance and algorithm governance, risk and problems of digital financial platforms, labor and user rights, antitrust and supervision, etc., jointly study the economic innovation and governance of the platform, provide the platform to provide the platform to the platform Policy suggestions and solutions for various types of economic problems make the platform economy more fair and fair, and then promote the orderly, stable and prosperous development of the economy.

- END -