From policy decision-making to theoretical research, there are two hottest terms related to data: one is data elements, which are relied upon to promote growth and provide new momentum. One is data property rights, hoping to promote transactions and achieve growth by defining data property rights. How should people define data property rights? What is the basis for defining data property rights? Is it necessary to define the property rights of data? The property rights economic theory pioneered by Coase, Alcin, and Demsetz discusses the property rights issues of industrial economic forms. In the form of industrial economy, resources are highly scarce and require systematic and legal property rights definition to solve the problem of effective allocation of resources. However, conducting research on data property rights using existing theories may be problematic. Many discussions about data property rights still follow the industrial economic paradigm of property rights, which means that regardless of the formation and characteristics of data, they either directly enter the welfare effect of data utilization or directly enter the issue of how to decentralize power. So how should we define data property rights? Let's first take a look at the characteristics of the data. So far, economists have formed the following three consensuses in their research on data characteristics. Firstly, the circulation and utilization of data may bring about complex externalities. On the one hand, as the scale and types of data elements grow, the knowledge and information formed by them will continue to accumulate, and positive externalities will rapidly increase, thereby exerting economies of scale. After reaching a certain scale, data elements have a promoting effect on the productivity improvement of all enterprises, and enterprises do not need to pay costs for this positive effect. On the other hand, the circulation and use of data may also bring about a series of negative externalities, such as privacy breaches, algorithmic discrimination, and data monopolies. The second is non competitive. This feature refers to multiple users or enterprises being able to use the same data resources simultaneously without affecting each other's usage. Any number of companies, workforce, or machine learning algorithms can use the same data simultaneously without reducing the amount of data used by anyone else. This non competitiveness will lead to increasing returns. The marginal cost of additional use of data elements is zero, which is a major difference between data and other production factors. The third is zero cost/low-cost replication. The cost of data elements mainly lies in the early stages of data acquisition and research and development, so the cost of initial data assets is high. However, the marginal cost of subsequent products tends to zero due to their infinite replicability. The development of digital technology has greatly reduced the variable costs associated with the transmission, replication, and use of data elements once they are formed, even approaching zero costs. The same data can be used simultaneously by different enterprises or entities. The low replication cost of data enables information to be quickly distributed and stored, enhancing instant access and availability of information, thereby promoting innovation and growth. The question is, if we follow the industrial economic form of property rights discussion, is it necessary to define property rights for data? Why should property rights be defined under the form of industrial economy? The problem that property rights need to solve is precisely due to externalities, competitiveness, and transaction costs. If economists have a correct understanding of the characteristics of data, it is necessary to discuss the necessity of defining property rights in data, especially because it has positive externalities in many contexts. The traditional property rights theory, which discusses the need to define property rights only because of externalities, no longer exists. Of course, property rights still need to be defined in situations where externalities exist. As for its non competitive nature, it also challenges the function of implementing property rights arrangements with restricted competition, and the solution to the problem of transaction costs brought about by zero or low costs. The necessity of defining property rights based on the existence of transaction costs in traditional theory is also challenged. It seems that either further research on the characteristics of data is needed to clarify its similarities and differences with the characteristics of factors in industrial economic forms, or another approach is needed to study the issue of data property rights. The formation, value realization, and property rights of data, however, we must not conclude that data does not require defining and enforcing property rights due to its uniqueness. How to find the path for defining and allocating data property rights is an important issue in data economics research. Looking back, the economic tradition established by Coase to study property rights from the perspective of transaction costs is worth reflecting on. They emphasized the importance of defining property rights for economic transactions and welfare effects, but overlooked the formation and value distribution of rights. If we follow this tradition to discuss property rights issues in the form of digital economy, it will either be aimless or reverse the order. We propose that discussing data property rights from the perspective of data formation and value appreciation may be a pathway. How does data transform from a resource into a data element? How do data elements ultimately become capital to achieve appreciation? 1. Data becomes an essential element. There are two aspects involved. Firstly, data is a byproduct of consumer activities and is generated by consumers. Data needs to become a factor that depends on the consumption of each consumer; The second is that data is recorded one by one into the database, becoming elements. 2. Data becomes a valuable element. Valuable data needs to go through the data production process, including data collection, storage, transmission, processing, and analysis. Whether it is the data used by data enterprises for transactions or the platform utilizing the value of data to achieve value-added, both must go through the production of data. 3. Value appreciation of data. Not all ways to increase the value of data are achieved through transactions. In addition to achieving value appreciation through data transactions, there is also a way for the platform center and platform enterprises to utilize the value appreciation of data, as well as the combination of data, computing power, and algorithms in the platform ecosystem. Therefore, to study the issue of data property rights, we cannot follow the industrial economic form of property rights economics and discuss data property rights issues from the perspective of data characteristics. Instead, we should discuss data property rights issues from the perspective of data formation and value appreciation. Here are several data property rights issues that need to be studied: firstly, individual data rights. After consumers generate data through consumption, do they only have privacy rights over the data elements formed? Do consumers have personal rights to the value generated by the use of data? This involves whether consumers only obtain their consumption needs through platforms or other organizations, or whether they also have a stake in the value of data rights; It also involves the issue of the rights of enterprises and platform hubs in the platform to use consumer data. Personal data rights are not only a meta issue of data property rights, but also involve data rights allocation and data governance issues that have long been overlooked by industrial economic property economics. The second is the rights structure and distribution of rights value in the formation of data value. The transformation of data from a resource to a factor, forming value and value appreciation, is a complex production process, including data recording, storage, processing, analysis, etc. Each process plays a role in the formation of data rights value, and the contribution of each link to the amount of rights is different, thus forming a complex data property rights structure. This involves both the division and measurement of data property rights value, as well as the allocation of data rights. The third is the distribution of platform data property rights and rights value. The core competitiveness of a platform is data-driven, without data, there is no platform's life. Platform data involves the consumption data of consumers at platform merchants, where merchants utilize the value added by consumption data, and the platform utilizes the value added by both merchant and consumer data. The property rights issues involved here include whether consumers have the right to claim the value generated by platform merchants and platform hubs using data? Do platform merchants have the right to claim the value generated by the platform's central utilization of data? The fourth is the combination of data elements and the value of rights. The realization and appreciation of data value depend on computing power and algorithms, but currently the combination of data, computing power, and algorithms is a black box. How are they combined? How do computing power and algorithms work? How does data form core elements through computing power and algorithms? This black box not only involves the construction of production functions, but also the issue of property rights, measurement of rights value, and distribution of data rights. (Xinhua News Agency) Author: Liu Shouying (Professor at the School of Economics, Renmin University of China, Secretary General of the National Digital Economy Professional Degree Graduate Education Guidance Committee)