[author]Li Jian

[content]

Digital platform regulation: rethinking publicness theory and applying economic regulation

*Author Li Jian

Professor of KoGuan School of Law, Shanghai Jiao Tong University

Adjunct Research Fellow of China Institute for Socio-Legal Studies, Shanghai Jiao Tong University

Abstract: The theory of publicness, represented by common carriers and new public utilities, hopes to change the situation of the ineffective implementation of antitrust laws, and to strengthen the constraints on the market power of digital platforms by granting special obligations. The theory of publicness has a long history, but the basic concept is vague and ambiguous, making it difficult to clarify the scope of regulation; the regulatory obligations advocated, such as non-discrimination and universal service, are difficult to be realized independently without the cooperation of measures such as structural divestiture, price control, and access restriction; and the implementation of regulation based on the nature of the platform's services will also result in excessive market intervention. The economic control with the kernel of natural monopoly theory is not only clearly defined, but also limits the scope of control to platforms with the nature of natural monopoly, which is more suitable for the characteristics of the rapid development of digital platforms. To realize the economic control of digital platforms, it is necessary to draw on the mature control theory and practical experience in traditional economic control, especially to structurally divest the functions of platforms with natural monopoly attributes and attach interoperability obligations to the divested natural monopoly platforms.

Keywords: digital platforms; common carriers; new utilities; economic regulation; interoperability obligations

1. Background and Issues

The rapid development of digital platforms has made them increasingly influential in social and economic life. Taking Apple as an example, its market capitalization exceeded $1 trillion for the first time in August 2018, $2 trillion in August 2020, and in January 2022, Apple once became the only company in the world with a market capitalization of more than $3 trillion. The market capitalization of Apple alone could rank 5th in the global ranking of countries' GDP in 2022, surpassing the UK's $2.67 trillion and trailing only Germany's $3.85 trillion. At the same time, digital platforms and people's daily lives are becoming more and more integrated. Whether in instant messaging, socializing, product purchasing, movie evaluation, or activities such as video games, travel, and online education, digital platforms play an important role.

Due to the characteristics of technology and business model, digital platforms have brought great impact to the established legal system. In particular, there is a heated debate on how to consider the legal system to continue to use or transformation (如何考虑法律体系的沿用与变革，)when facing the market power of digital platforms. Antitrust law is the most prominent law that imposes restrictions on market power, and research around its value falls into two main categories: one type of research focuses on the role of antitrust law, but there are different perceptions. Some argue that antitrust law should be changed to a greater extent in order to change its insufficient implementation, while the opposing view is that antitrust law does not need to be changed and that existing rules can provide more room for the market. Another group of studies argues that the nature of antitrust law makes it difficult to effectively regulate digital platforms, and that there is a need to look beyond antitrust law and focus on the impact of the platform economy on society as a whole, and then seek special regulation of digital platforms. This kind of viewpoint with "publicness" （公共性）as its core has generated greater academic and policy influence. Among them, the most representative are the common carrier theory and the new public utility theory (hereinafter referred to as the "publicness theory")（公共性理论）. The common carrier theory advocates treating digital platforms as common carriers and requiring them to be bound by the principle of neutrality. The New Utility Theory emphasizes the increasingly fundamental role of platforms in the modern economy and advocates treating digital platforms as a new type of public utility.

As they are also faced with the question of how to effectively regulate digital platforms, the thinking of foreign scholars in recognizing platform regulation in terms of publicness has also influenced the views of domestic scholars. Before 2021, there was not a single case of antitrust enforcement against digital platforms in China, and although the "antitrust storm" of late 2021 resulted in a number of cases of monopolization of platforms, the technical difficulties of antitrust analysis that it revealed have deepened questions about the role of antitrust law. Therefore, domestic scholars are generally dissatisfied with the effect of antitrust enforcement in the field of digital platforms. With the help of public nature theories such as common carrier and new public utility, it is undoubtedly very attractive to construct a whole set of ex ante control measures in addition to the antitrust law characterized by ex post control, which not only avoids complicated antitrust technical analysis, but also imposes clearer and more direct restriction on the operation of platforms. In this sense, the theory of publicity responds powerfully to the reality of the problem. In fact, the Digital Marketplace Law adopted by the European Union in 2022 and the Guidelines for the Classification and Rating of Internet Platforms (Draft for Public Comments) issued by China's State Administration for Market Regulation in 2021 (the "Platform Guidelines") are products of these theoretical perspectives. Among them, the EU's Digital Markets Act, which was enacted earlier and has systemic integrity, has had a more direct impact on China at the practical level.

Publicity theory involves the basic logic of platform regulation, and therefore needs to be analyzed under the framework of regulation theory. Regulation usually includes antitrust law (regulation), economic regulation, and social regulation. Among them, antitrust law focuses on how to maintain the competition mechanism so that it can play the role of resource allocation; economic regulation mainly applies to industries with natural monopoly attributes, such as water, power supply, telecommunication, postal service, railroads, etc., and restrains the operators through access restrictions and price limitations, etc.; and social regulation focuses on the fields of environment, safety and health, etc., and solves the problems brought by information asymmetry and externalities. In terms of restraining the economic power of operators, the difference between antitrust law and economic regulation lies in whether a particular industry or market is competitive: if the market is competitive and can enhance efficiency through competition, then antitrust law will be applied to regulate the market; if the market is not competitive, which means that competition will bring serious inefficiency problems, then the government will intervene through economic regulation. Antitrust law and economic regulation form a complementary relationship to deal with market power constraints. However, the publicness theory of digital platforms breaks this basic theoretical understanding. Although the theory of publicness points to the insufficient implementation of antitrust law, it does not resort to economic regulation after jumping out of antitrust law, but directly builds a new regulatory system from scratch. Regulation based on the publicness theory emphasizes the public attributes of the products or services provided by the relevant platforms rather than whether the market is contestable or not. As such, there are significant differences in the foundations of the publicness theory compared to the established regulatory theories.

For the regulation of digital platforms, the established theoretical divisions of regulation remain valuable. Although from a practical point of view, the boundary between antitrust law and economic regulation is not clear, and there have been a number of antitrust cases involving regulated industries in both the United States and China. Some of the new regulations, including the EU Digital Markets Act 2022 and the German Law against Restrictions of Competition 2021, are similarly vague in characterization. Instead of following the basic framework of antitrust law and economic regulation, the EU Digital Marketplace Act directly identifies "core platform operators" by turnover, market capitalization, number of active users, and scope of business, and imposes common antitrust obligations such as the prohibition of restraint of trade and tying. Section 19a of the German Anti-Restriction of Competition Act newly introduces the concept of "operator with significant cross-market competition effects", which is a new standard in addition to the traditional market dominance determination, in which the determination conditions such as "dominant position in one or more markets" are different from the existing antitrust law theoretical framework. The concept of "dominant position in one or more markets" is a new standard in addition to the traditional market dominance determination. However, the misalignment between theory and reality does not mean that the established theory has lost its significance. The reasons for the lack of strict distinction between economic regulation and antitrust law include both a lack of theoretical understanding and historical inertia. The former led to an overly direct response to the needs of reality, and the latter is reflected in the path dependence on prior jurisprudence developed after the early antitrust law intervened in the regulated industry. In fact, whether it is the misallocation of legal resources brought about in individual cases or the ambiguity of the nature of digital market law, the fact that these practices have been questioned does not negate the rationality of the original theory. Here is not to insist that the original theoretical cognition is unchangeable, but to emphasize that, in the face of the impact brought by the theory of the public nature of digital platforms, it is more crucial to examine the inherent reasonableness of its theoretical logic in conjunction with the established theoretical framework. And a good theoretical framework is also the basis for judging whether the practice is reasonable or not. In this sense, the proof of the theory of publicity still faces great challenges.

2.The ambiguous definition of the theory of publicity

The theory of publicity mainly summarizes and refines the conceptual system by combing the jurisprudence of various historical periods, and the jurisprudence resources it draws on are relatively complex, including public utilities, common carriers, and net neutrality, etc. These concepts have a long history and are often used as a basis for judging practices. These concepts are characterized by a long history of fluctuating connotations. In the relevant research literature, these concepts are sometimes specifically distinguished and sometimes used as interchangeable terms, which has led to a great deal of confusion. A proper sorting out of the concepts is undoubtedly conducive to a better understanding of their connotations and provides a basis for judging the reasonableness of the relevant regulatory measures. Generally speaking, the terms related to the theory of publicity traced from case law are vague and lack a commonly recognized definition.

2.1 Three Views on the Recognition of Common Carrier

The concept of common carrier can be traced back to England in the 15th century, and gradually developed through the common law. In fact, it is not easy to sort out a clear path of institutional evolution from the long and complicated common law. According to the current consensus, common carrier originated from the English common law concept of "public call", which refers to the commitment to serve the public, including innkeepers, blacksmiths and surgeons and other professions include such a commitment. Thus, service to the public without discrimination was recognized as a fundamental obligation of the common carrier. Accordingly, the identification of a common carrier becomes the basis for determining the reasonableness of government regulation. In other words, if the operator constitutes a common carrier, the government can regulate to ensure the fulfillment of the above obligation. However, there is no consensus on how to clearly define common carrier, whether from court decisions or academic research. Looking back at the history of development, there are at least the following three mainstream views on the identification of common carriers:

The first view is that a common carrier is an operator with a monopoly position, and the nature of its obligations is a special obligation to be borne by the monopolist. However, as a purely factual matter, a number of subjects lacking monopoly power have also been found by the courts to be common carriers, such as inns, trucking companies, and affiliated long-distance railroads. In these cases, the correlation between market monopoly power and the nondiscrimination obligation is minimal, and the finding of common carrier status itself constitutes the source of the nondiscrimination obligation.

The second view asserts that an operator is a common carrier as long as he or she himself or herself claims to be open to the public. For example, the United States Court of Appeals for the D.C. Circuit, in National Association of Regulatory Utility Commissioners v. Federal Communications Commission, asserted that a common carrier is an operator that is committed to making its services available to all and "markets itself without reservation to customers for whom it is suited". However, allowing an operator to determine whether it is a common carrier based on a description of its own services necessarily introduces manipulation and uncertainty. All a provider needs to do to avoid common carrier obligations is to limit the services it provides to a portion of its entire customer base, rather than the public at large.

A third view asserts that common carriers apply to industries "affected by the public interest". This approach was most famously articulated in the landmark case of Munn v. Illinois ("Munn"). In that case, the Court held that industries including ferries, wharves, warehouses, taverns, hotels, mills, bridges, highways, and common carriers were "affected by the public interest." The idea that the public interest is affected in order to find digital platforms subject to special regulation is also central to current theories of publicity. It is important to note, however, that this view is also controversial. For example, in the Munn case, Justice Field argued that the public has an interest in industries as diverse as housing, textile manufacturing, machine building, and book printing. The United States Supreme Court in Nebbia v. State of New York (hereinafter referred to as "Nebbia") also recognized that there is no separate test category "affected by the public interest". In Jackson v. Metropolitan Edison Company, the United States Supreme Court also rejected the proposition that industries "affected by the public interest", such as electric utilities, have a state role, citing Nebbia, and holding that "affected by the public interest" is "undefinable and an unsatisfactory test".

Although there is no consensus on how to define a common carrier, it is generally recognized in the academic community that, once identified as a common carrier, the relevant operator is subject to an obligation of non-discrimination, i.e., to treat its customers without discrimination. Such non-discrimination requires that the common carrier shall provide its services on fair and reasonable terms and conditions, without any unjust or unreasonable discrimination or any undue or unreasonable advantage to any particular person or class of persons. For example, railroads recognized as common carriers are required to provide standard service to the public and should not differ in pricing or in the services they provide to their customers. Overall, these obligations of a common carrier depend on the type or function of the service it provides and are not related to the structure of the market (i.e., the number of service providers).

2.2 Three Drivers of the New Utility Theory

The New Utility Theory was developed based on the understanding of traditional utilities and is used to address the theoretical underpinnings of platform regulation. Therefore, in understanding the claims of the New Utility Theory, one needs to first identify what a utility is. Historically, the development of utility theory has been driven by three factors:

The first was dissatisfaction with the railroad industry. The advantages of railroads over water and road transportation, reinforced by government involvement in planning and designing routes, constructing rights of way, granting concessions, and providing land, resulted in railroad companies facing little competition. As a result, railroads were able to use market forces to engage in discriminatory practices in pricing and service. In response to public outcry, the United States eventually established an independent regulatory agency, the Interstate Commerce Commission, to regulate prices and services in the railroad industry. Railroads became the first industry to be classified as a public utility.

The second is the "affected by public interest" standard proposed by the Court. In Wolfe-Parkin Co. v. Kansas Industrial Relations Court (hereinafter referred to as the "Parkin Co. case"), the United States Court suggested that if a product affected the public interest, it would require special regulation. At various times, the Court has included products and services that do not have market power, such as cotton presses, news media, tobacco warehouses, insurance and ice for sale. There is certainly a need to interpret "affected by the public interest", but unfortunately, these interpretations are not uniform. In the Parkin case, Chief Justice Taft identified three types of "affected by the public interest": (1) companies that operate under privileges granted by the government, which either explicitly or implicitly impose a duty to provide public services. (2) Certain special occupations whose impact on the public interest was recognized long ago and survived the period when Parliament or colonial legislatures arbitrarily enacted laws to regulate all trades and occupations. These occupations included the hotel trade, the cab business, and the millinery trade. (3) Though not public in their inception, the proprietors put them to public use, thus enabling the public to derive a benefit from that use, and subjecting themselves to public regulation to the extent of that benefit.

Third is the acceptance of the idea of natural monopoly. John·Stuart·Mill was the first economist to talk about natural monopolies when he observed that certain public services in London could not be supplied competitively.

The third is the acceptance of the idea of natural monopolies. John Stuart Mill was the first economist to talk about natural monopolies when he observed that certain public services in London could not be supplied competitively. Thomas Farrer made the first attempt to identify natural monopolies by their economic characteristics in 1902, when he categorized industries that had never tried to compete, or that had tried to compete but failed, as natural monopolies. These ideas have influenced the way society views natural monopolies. For example, the United States Securities and Exchange Commission applied the concept of natural monopoly to energy utilities in the Public Utility Holding Company Act of 1935.

It is the multiple factors that have influenced the development of the concept of public utility that have made it difficult to have a single definition that can unify the different stages of understanding of public utilities. Indeed, many descriptions of the concept are circular. For example, "A utility is a business that provides daily necessities to the public at large. Utilities provide water, electricity, natural gas, telephone service, and other necessities." What is more interesting, however, is that scholars advocating for a new utility theory tend to cite the uncertainty of the utility concept as an advantage. They argue that the utility framework is surprisingly flexible, with applications ranging from electricity and transportation to goods and services such as banking, milk, ice, and water, and that the categories of goods included in it change as technological and economic conditions change. Accordingly, a growing number of legal scholars have linked utilities to areas such as finance, energy, and pharmaceuticals as a way of diagnosing and solving the problems of modern infrastructure regulation. Objectively, however, the advantages of the so-called "flexible framework" are due more to a lack of conceptual ambiguity than anything else.

Despite the lack of a unified understanding of public utilities in the legal community, economics has used natural monopoly theory to analyze public utilities. However, the new utility theory does not resort to this path of understanding, but rather offers its own interpretation of the history of the development of the concept of utility and applies it to digital platforms. The proposition can be summarized in two core aspects: first, utilities provide products and services that constitute "necessities". Second, when companies gain control over the provision of and access to necessities, stronger regulation is needed. In this understanding, the new utility emphasizes those products or services that affect the public interest by virtue of their necessity and thus exhibit scale, necessity, and vulnerability (e.g., Internet access services, finance, information platforms). Therefore, by selecting the dimension of "affected by the public interest", the new public utility theory argues that digital platforms constitute public utilities in this sense, and thus require special regulation.

2.3 Reflection on the Concept of Publicity Theory

Through a brief overview of the development of the concepts of common carrier and public utility, the following conclusions can be drawn:

1. there is a great deal of overlap between the concepts of common carrier and public utility. The distinction between common carrier and public utility is not clear in the academic literature. For example, the Munn case of 1876 is sometimes cited in different literatures as an important case for common carriers and sometimes as a key case for public utilities. Similarly, some operators are considered both common carriers and public utilities, such as railroads and telephone companies, while others are considered only public utilities, such as gas and electric companies. Although the boundaries are not so certain, some scholars have argued that common carrier is a much more inclusive concept that applies not only to public utilities, but can be applied to utilities other than public utilities. Regardless, this conceptual overlap, while reflecting the common concern of common carrier and utility theories - the issue of publicness - also suggests that the theoretical streams are not clear and can lead to confusion in the discussion; after all, the two are significantly different in terms of, for example, regulatory obligations, such as common carriers focusing more on non-discrimination obligations, while utilities more emphasis on the affordability obligation.

2. Lack of Consistent and Clear Definition. It is difficult to obtain a consistent framework or definition of common carriers when they are generalized entirely from a historical perspective. The same is true of the utility concept. Much of this inconsistency is related to the research methodology employed by the theory of publicity. Publicity theory seeks a rational basis for current theoretical claims primarily by distilling legal principles and combing case law. But the longer the history, the more difficult it is to refine concepts as a result of political, economic, and cultural changes. In order to accomplish this task, concepts are necessarily made as vague as possible to improve coverage. For example, scholars of publicness theory have redefined the category of firms "affected by the public good" as "infrastructure", i.e. resources that create positive externalities. However, this does not solve the problem of the difficulty of defining "affected by the public interest" and has led to different interpretations in the aforementioned cases. In the absence of a clear definition of "affected by public interest", it is of course difficult for a new definition to present consistency and clarity. Therefore, the theory of publicity has been questioned historically. For example, it has been argued that far from being a tried and tested regulatory solution, common carriers are vulnerable to inefficiencies, structural biases, and manipulation. This will undoubtedly limit the application of the theory of publicness.

3. avoidance of the monopoly status element. In the history of the development of public carrier and utility theory, the definition through the monopoly position of the operator is one of the important points, but the current theory of publicity seldom mentions it. Take the landmark Munn case as an example. The case involved grain silos in Illinois that, according to the United States Supreme Court, were uniquely positioned between river ports and railroad tracks to control the transportation of grain from farmers in certain Midwestern states to East Coast markets. Since it was virtually impossible to move the ports or railroad tracks, the silos had a "virtual monopoly" on the storage and transportation of grain from seven or eight large Midwestern states. The Court ultimately concluded that it was appropriate for the government to use its power to control the conduct of these enterprises because their actions had such broad consequences that they were affected by the public interest. In this case, if the grain silos did have the status found by the Court, they constituted a monopoly and could even be "core facilities" for purposes of the antitrust laws and thus be subject to facility-sharing obligations. In fact, the public interest is affected by emphasizing the uniqueness of the product in question and thus making it a necessity, as reflected in other cases. It is therefore not appropriate to ignore the monopoly position. Of course, as noted earlier, the publicity theory would like to strengthen the regulation of digital platforms outside of the antitrust laws. In order to avoid pulling the theoretical line of sight back to the antitrust laws, downplaying or excluding the monopoly factor from the historical lineage becomes an inevitable practice in the claims of publicity theory.

By searching for historical resources and thus seeking a theoretical basis for contemporary regulation of digital platforms, publicity theory does not provide a clear definition to better determine what constitutes a common carrier and how it constitutes a new utility. This lack of clarity inevitably leads to different understandings of the same issue. For example, in social media-related litigation, at almost the same time, the U.S. Court of Appeals for the Eleventh Circuit held that platforms were not common carriers in Knight Joyce, LLC v. Attorney General of Florida, but the U.S. Court of Appeals for the Fifth Circuit held that platforms were common carriers in Knight Joyce, LLC v. Pancoaston. When the definition is not clear enough, reviving it or repackaging it for digital platform regulation is just converting one vague issue into another. Of course, this is not to deny that digital platforms have a certain degree of public nature, but to emphasize that direct regulation based entirely on the vague "public interest" will make the theory lack of a sound foundation.

3. Regulatory measures of publicness theory and their limitations

Rather than simply proposing new concepts, the theory of publicity draws on these concepts in the service of eventual regulatory measures. For example, the approach to utility regulation in digital markets proposed by Rahman, an advocate of the new public utility, includes very specific measures: first, identifying the products and services that serve as "infrastructure". Second, address the problem of digital infrastructure overreach through three sets of tools: regulation by agencies such as the Federal Trade Commission (FTC) or the Federal Communications Commission (FCC), firewalls or structural separation between the products and services that serve as the infrastructure and others provided by the same entity, and public/private choice through nationalization, municipalization, and competition between the government and other firms in a given market segment. Choice. Despite the lack of clarity on the scope of common carriers and new utilities, the policy directions of the publicness theory are relatively clear.

The theory of publicness contains a mixed bag of regulatory policies, but it focuses on two regulatory claims: for common carriers, regulation centers on the requirement that these firms treat all customers equally; for utilities, regulation is that because the service is critical, the government must ensure that the service is affordable. This distinction is a good abstraction of the two regulatory obligations that are most fundamental to the theory of publicity, namely "non-discrimination" and "universal service". By analyzing these two aspects, one can better see the policy implications of the public nature doctrine, as well as the relevance and possible inherent conflicts between different regimes.

Non-discrimination and universal service obligations are core requirements of the common carrier, new public utility doctrine, but often cannot be implemented in isolation as a means of direct regulation. Typically, once recognized as a "common carrier", the operator is subject to four main obligations: access restrictions and service obligations, the obligation to charge non-discriminatory rates, the obligation to charge just and reasonable rates, and structural unbundling. These requirements are widely reflected in various laws and regulations as well as judicial cases, such as the rule against unreasonable discrimination in the U.S. Federal Communications Commission's 2010 Preserving the Open Internet: Broadband Industry Practices (Open Internet Order), the 2015 Protecting and Promoting the Open Internet (2015 Open Internet Order) in which lawful content cannot be blocked, and traffic cannot be controlled based on content, services, etc. , and general standards of conduct such as the prohibition on paid prioritization. Similar to the many obligations of public carriers, the regulatory measures and obligations attached to utilities typically include: controlling market entry or exit; establishing minimum conditions for service and quality control and other consumer protections; and universal service obligations, i.e., the obligation to serve all customers within a well-defined geographic area, including the obligation to provide service at prescribed "just and reasonable" rates. The requirement to provide services at prescribed "just and reasonable" rates. The reason for the multiplicity of regulatory measures is that the core obligations hardly stand alone and need to be complemented by other measures in order to be truly effective. These, in turn, are ignored by proponents of the publicness theory.

3.1 Non-discrimination obligation and price control, vertical integration

The essence of discriminatory behavior lies in selling the same product or service to different customers at different prices. Therefore, whether or not an operator has engaged in price discrimination needs to be determined by whether the products or services are "identical" and whether there are "different" prices, rather than simply by whether the final prices are the same. In addition, there is a need to prevent operators from substantially discriminating in price by altering the quality of the product when the price is apparently the same. However, careful differentiation necessarily entails significant enforcement costs, contrary to the certainty and low cost required for direct regulatory instruments. License fees for standard-essential patents are a good example. The licensing of standard-essential patents needs to satisfy the principle of "fair, reasonable and non-discriminatory" (FRAND), but because the standard-essential patent organization does not directly set the licensing conditions of the patent, resulting in the actual licensing of what is "fair, reasonable and non-discriminatory" price disagreement, and has led to the emergence of the "fair, reasonable and non-discriminatory" price of the licensing of the standard-essential patents. However, the fact that the SOPO does not directly set the licensing conditions of patents has led to disagreement on what is "fair, reasonable and non-discriminatory" in actual licensing, and a large number of rate disputes. Therefore, when implemented as a direct regulatory tool, the implementation of the non-discrimination obligation often implies the setting of specific sales prices, so as to facilitate the enforcement agency to directly and explicitly determine the legality of the act. For example, the natural gas price in Shanghai was adjusted in 2021, directly specifying different prices for different user categories. Then, when regulation is achieved in a low-cost manner, even for digital platforms, the price of the relevant product or service needs to be set. Since the publicity theory is based on the nature of digital platforms, without distinguishing between types of platforms, and the non-discrimination obligation it advocates is likewise a universal requirement for all digital platforms, it necessarily involves a wide range of product or service categories, which makes price-setting highly challenging.

There is also a close correlation between the realization of the non-discrimination obligation and the way in which the operator is organized. Typically, if the overall price of a product or service is not set, the non-discrimination obligation makes it difficult to prevent vertically integrated firms from excluding firms that are unaffiliated and offer complementary products or services. A vertically integrated firm could simply charge the same high price to upstream and downstream non-integrated firms, and a non-discriminatory high price would have the same exclusionary effect on those firms as a discriminatory price. Thus, in the absence of structural divestitures, actual nondiscrimination provisions would fail to restrict profit shifting within vertically integrated firms. There is a large body of empirical research in this area. And the most important way to change this result would be to separate the relevant businesses, such as splitting an electric utility's generation, transmission, and distribution businesses into three separate companies. However, structural divestitures have very significant effects on market competition and may result in efficiency losses. This is because vertical integration allows the operators involved to internalize the benefits they create to mitigate the underproduction associated with positive externalities or can reduce prices, especially if both upstream and downstream markets are highly concentrated. This is well recognized in economics. In the platform economy, it is very common for digital platforms to compete across borders by leveraging the characteristics of bilateral markets. However, it is undeniable that structural divestiture is the most economical and reliable way to implement platforms' non-discrimination obligations in direct regulation. If such a structural solution is not adopted, it would mean a case-by-case examination of whether the product or service has fulfilled the non-discrimination obligation, which is not only contrary to the approach of direct regulation, but also incurs high enforcement costs. There is certainly a question of how to strike a balance here, but this is not considered by proponents of the publicity theory.

3.2 Universal service obligations and cross-subsidization

The obligation to provide affordable services is linked to universal service. Universal service means that in order to satisfy the public interest, operators should provide basic, affordable services to all. The obligation exists in areas such as basic telephone services, Internet services and broadband services. Universal service usually encompasses two reasons: one is to redistribute consumers who do need the service in question and to protect them from the adverse effects of large price increases. Target groups include low-income residents, persons with disabilities, the elderly and rural consumers with limited access to transportation. The second is regional development planning, which encourages residents to migrate from congested large cities, leading to a more harmonious population distribution. Universal service inherently includes cross-subsidization. Cross-subsidization, which cannot be avoided in universal services, exists when firms offering more than two products or services use revenues from one to compensate for the additional cost of providing another, or when money earned with one group of consumers is used to subsidize another. For example, regulated prices for local telephone services in rural areas often do not adequately reflect the relatively high unit costs of providing services to these areas, even though prices in rural areas tend to be close to (and sometimes below) costs, while prices in urban areas tend to be considerably in excess of costs. Only such a pricing model can ensure that basic telephone services are affordable for all citizens.

3.2.1 Access restrictions and monopoly profits

Market access restrictions are necessary to ensure that cross-subsidization can take place and thus universal service obligations. For cross-subsidization to take place, there is a need to ensure higher profits for a subset of products or services that would otherwise not be subsidized. If these products or services are exposed to competition, this inevitably leads to lower profits. Therefore, profits can only be ensured, and thus subsidized, if there are regulatory barriers to entry. To better illustrate this, suppose that Alibaba Group's e-commerce platform is subject to universal service obligations in Tibet and Xinjiang. Because of the large population size and high density in Jiangsu, Zhejiang and Shanghai, distribution costs are low, while in Tibet and Xinjiang, because of their sparsely populated areas, distribution services are bound to be more costly, and ultimately the price of the service is high. To make e-commerce services affordable, it is necessary to subsidize Tibet and Xinjiang to reduce the price of services. However, the premise of the subsidy is that Alibaba Group can make higher profits at lower costs in other regions-Jiangsu, Zhejiang, and Shanghai. Moreover, taking on universal service would cause an increase in Alibaba Group's overall costs. Without market entry restrictions, competitors would enter the Jiangsu, Zhejiang, and Shanghai regions, lowering service prices through competition and compressing Alibaba Group's profit margins. Especially when the entrant is not required to assume the universal service obligation, its cost will be significantly lower than that of Alibaba Group. Thus, the entry of new entrants into a market that is fully open to competition induces a "cream-skimming" of profitable segments of the market, making it impossible for operators with universal service obligations to finance those obligations through cross-subsidization.

The essence of the problem would not change if universal service obligations were imposed on every operator entering the market. The limitation here lies mainly in how much profit a high-margin market can shift to subsidize high-cost areas or consumer groups. When market capacity is certain, more entrants imply higher output and lower profits, with a corresponding reduction in the profits that can be subsidized, and lead to a reduced ability to afford universal service obligations. As a result, in these markets, the market structure is usually exclusive or oligopolistic. This also means that for digital platforms to be able to afford universal service obligations, it is proper to maintain the monopolistic position of the platforms.

3.2.2 Affordability and price discrimination

Affordability of a product or service also raises the issue of discrimination, as affordability is related to differences in the ability to pay of the group in question. The same $2,000 cell phone does not pose a financial burden for affluent families, but is difficult for poor families to pay. The affordability problem can be solved by charging $3,500 to affluent families and $500 to poor families. However, this also brings up the issue of price discrimination and has triggered much controversy in the platform economy. For example, in the "big data ripeness" discussion, the behavior of charging high prices to highly dependent consumers is usually given a negative evaluation. However, the platform's "ripening" behavior is usually accompanied by a large number of discounts for new customers, i.e., "benefiting" behavior. In other words, the "affordability" of new customers comes at the expense of existing customers, who are usually more dependent. Thus, there is an irreconcilable conflict between non-discrimination and affordability for products or services that are widely desired and have the capacity to affect the public interest. The subsidies required for affordability come either from cross-subsidization of other products or services or from direct government subsidies. In a platform economy, it is inherently unattainable to simply require reasonable, non-discriminatory rates without regard to the other.

3.2.3 Universal regulation and platform categorization

The theory of universality emphasizes obligations such as non-discriminatory and affordable obligations for platforms based on the nature of the product or service they provide. However, such a requirement of universality would make the scope of regulation very broad, which would not only result in excessive intervention in the market, but also make regulation very costly. Therefore, in reality, regulatory measures are often limited to specific areas and platforms of specific sizes. For example, the "core platform services" stipulated in Article 2 of the EU Digital Market Law include online intermediary services, online search engines, online social networking services, video sharing platform services, number-independent interpersonal communication services, operating systems, cloud computing services, and online advertising. The types of platforms stipulated in Article 2(2) of the "Guidelines" include six categories: online sales platforms, life service platforms, social entertainment platforms, information platforms, financial service platforms and computing application platforms. Comparatively speaking, the types of platforms in China's "Platform Guidelines" are more extensive, covering the main parts of products or services provided through Internet platforms, and requirements on the scale of platforms are also set out (Article 3).

Although the regulatory approach of "service category + scale" adopted by China can be targeted to require some platforms to undertake special obligations, this regulatory idea lacks a rational basis. Not to mention why these service categories constitute the core platform services, the scale of different industries (services) itself varies greatly. For example, the "Platform Guide" contains three categories of online sales platforms. Taking the comprehensive commodity trading platform as an example, it refers to a comprehensive platform that specializes in or is mainly engaged in providing various types of commodities, such as clothes, hats, shoes and boots, bags and accessories, digital appliances, food and toiletries, and so on. The market capitalization of Alibaba, Pinduoduo and Jingdong, which belong to this platform, is very impressive. In contrast, the overall scale of some industries is on the small side. For example, in the search engine class platform, the domestic market share of more than 70% of the Baidu company, its market value is also difficult to match the Jingdong Group. If search engine companies are excluded from the scope of special regulation because of their small market capitalization, there is an irreconcilability with the public nature theory that emphasizes the nature of the product or service as the basis for regulation.

More precisely, there is no direct correlation between market size and the existence of a strong market position, nor is there a direct correlation between the impact on the public interest - which can be ensured by market discipline if there is competition. "Article 3(3) of the Platform Guidelines emphasizes that super-platforms have a "super-restrictive capacity, i.e., the platform has a super-restrictive capacity to limit the access of merchants to consumers (users)". This is contrary to the rationale of antitrust law. Antitrust analysis usually involves defining the relevant market for the purpose of determining the extent to which competition occurs, which is often not directly related to the overall size of the market. For example, in the case of abuse of dominant market position by Eastman (China) Investment Management Co., Ltd. investigated by China's anti-monopoly law enforcement agency, the relevant product market was the market for alcohol ester twelve film-forming auxiliaries, and the annual sales of the dominant operator were less than RMB 500 million. In the case of Shanghai Food Pace Business Development Co., Ltd.'s administrative penalty for monopolization in the market for Internet catering delivery platform services, the annual sales of the dominant operator were even less than RMB 40 million. Even water and natural gas companies, which are usually public utilities, may not have high sales in a particular city due to geographical constraints, but their market position in the relevant market is so high that price control is required.

In summary, it can be seen that the regulatory measures recommended by the public nature doctrine largely lack an understanding of the interconnectedness of the regulatory instruments, rendering these policy recommendations unsystematic and unworkable, and inherently conflicting with their claimed ability to enhance the competitive effects of markets. Indeed, it is not only non-discrimination and universal service obligations that suffer from these problems, but also a number of the recommendations of the publicness theory are oversimplified, with little regard for the way in which they are to be implemented and the changes that they will bring about in market competition, making them difficult to respond to in further questioning. For example, the new public utility theory argues for the establishment of a specialized regulator to prevent platforms from unduly interfering with traffic. However, it is difficult for the theory of public utility to provide additional theoretical resources when these questions are not well resolved as to how to judge the "undue" interference of platforms with traffic, and whether platforms are likely to interfere with traffic inappropriately. Here we are not demanding that a foundational theory should provide all the specific measures and include all the details, but emphasizing the responsiveness of the theory itself. That is to say, specific systems can be appropriately deduced from this theoretical framework, or good guidelines can be provided for specific systems to support further refinement. In turn, when applied, the relevant theory can withstand the test of real-world problems and has logical consistency.

4. Regulation of digital platforms based on economic regulation

In its early development, the Internet was portrayed as an autonomous, utopian sphere of free social interaction, free from external interference from the state or the law, and as an autonomous space separate from the nation-state. However, the Internet has never been an autonomous, utopian sphere of freedom. Lawrence Lessig's idea that "code is law" is undoubtedly the best illustration of this. State regulation of market operations is inevitable; the more complex question is how this can best be achieved. As noted earlier, among the regulatory approaches to constraining market forces, antitrust law, which promotes competition and thus economic efficiency, is based on the competitiveness of the market, while economic regulation, which takes the uncompetitiveness of the market as its starting point, constrains corporate behavior through government intervention. Compared with the theory of publicity, the economic control based on natural monopoly has obvious advantages, and can absorb the core concepts of the theory of publicity to better realize the limited control of digital platforms.

4.1 Applicability of economic regulation

As a way of direct control, economic control is centered on the theory of natural monopoly, and cost disadvantage is the basis for defining natural monopoly. Simply put, cost-inferiority means that at a given level of output, it is less costly for one firm to produce it than for multiple firms to produce it in total, regardless of how output is distributed among multiple firms. Economies of scale are closely associated with cost-inferiority. Regulated industries based on natural monopolies usually have significant incremental returns to scale and substantial sunk costs that account for a large proportion of total costs, representative of such industries as water, landline telephony, and electricity grids. The existence of natural monopolies means that markets are usually limited to one or a small number of operators in order to maintain an efficient mode of production, and such market structures, in turn, inevitably lead to the abuse of this market position by profit-maximizing operators, resulting in harm to the interests of consumers or other operators. It is through direct restrictions on market entry, prices and other market conditions that economic regulation ensures the efficient operation of the market while at the same time curbing the abuse of market power by operators.

Economic regulation centered on natural monopolies is based on the fact that markets are not contestable, and targets the inefficiencies that can result from market competition. Some basic features of digital platforms fit well with the theory of natural monopoly. First, digital platforms have the attributes of a bilateral market. Bilateral markets have network externalities between different groups of users, and if the network externalities can be increased within the demand range, the platform may have natural monopoly. This is because when more users connect to the platform, even if this brings about a rise in the platform's operating costs, as long as the benefits of network externalities outweigh the costs, it may still be more efficient for one platform to provide the service. Second, digital platforms have significant efficiencies of scale. Similar to traditional utilities and network industries, the investment costs of network construction and infrastructure required for platform operations are high, but will fall significantly when outputs increase. Compared to traditional physical platforms, technological change has strengthened the trend toward concentration, and the digitization of platforms has also increased the potential for large supply-side economies of scale at low or zero marginal cost and expanded the range of matchmaking capabilities for users benefiting from network externalities, and for data collection and analysis to create important economies of scale and scope. The combination of these factors makes it entirely possible for digital platforms to constitute natural monopolies. In reality, we are also able to observe the trend towards concentration in the digital economy and the persistent monopoly position of some operators in niche markets.

It is important to emphasize here that not all digital platforms should be deemed to have the attributes of a natural monopoly. The types of products or services involved in digital platforms are very diverse, and the differences in economic attributes between different products or services are significant, even in terms of the bilateral market characteristics of digital platforms. Bilateral markets can be categorized into a variety of types, and network externalities are manifested in different ways, with differences between cross-network externalities and unidirectional network externalities, as well as between intra-group network externalities and inter-group network externalities. The existence of these differences not only indicates the diversity of digital platforms themselves, but also inevitably makes it different whether different types of digital platforms have natural monopoly properties. Although current digital platforms increase the likelihood of natural monopoly in general by processing data to achieve matching and prediction functions and by leveraging digital technology to embody scale efficiencies to a greater extent than previous bilateral market platforms such as dating clubs and credit cards, the degree to which different types of platforms exhibit it still varies and cannot be generalized. Herbert Hovenkamp, for example, argues that social media platforms are not natural monopolies because "in a stable natural monopoly market, a dominant firm need only charge a competitive price to exclude rivals or to fend off the occasional attack from others; in the absence of exclusionary practices, the market itself will determine how many firms remain in the in the market. The persistence of exclusionary practices thus shows that this market is not a natural monopoly." Francesco Ducey further divides digital platforms into three categories based on strength on natural monopoly attributes: first, high natural monopoly platforms, represented by horizontal search. Search engines have high fixed costs and very low supply-side marginal costs, with economies of scale similar to those of the standard web industry, and the economies of scale and scope from larger data sets are crucial for the purpose of improving the predictions of search algorithms. Second, low or no natural monopoly attribute platforms, represented by e-commerce platforms. While the online shopping market benefits from positive network externalities for both buyers and sellers, product differentiation in turn represents an important countervailing force to natural monopoly. Ride-hailing platforms, on the other hand, lie somewhere in between, as the value of network externalities and the ability to improve matching diminishes after a critical mass of users is reached, and the supply-side costs of entry are not particularly high. Such a typology of ideas allows for a more focused analysis of whether particular digital platforms constitute natural monopolies.

4.2 Advantages of economic regulation

Economic regulation is logically consistent with the theory of publicness. The public nature theory focuses on whether the relevant product or service constitutes a necessity of life, and whether it affects the public interest as a result. Products or services with the attribute of natural monopoly are indispensable to people's daily life, and will undoubtedly affect the "public interest". Therefore, the economic control centered on the theory of natural monopoly can be fully applied to the control of digital platforms, and has the advantages of clear definition and limited intervention.

1. Clearer definition. The biggest problem of the theory of publicity is unclear definition, "affecting the public interest" in the public interest has always been difficult to have a clearer definition, which causes great trouble to the application of the theory. In contrast, in the theory of natural monopoly relied on by economic regulation, the definition of natural monopoly based on cost disadvantage is not only clear, but also has the basis of quantitative analysis, and can be more certain to use the tools of economics to make judgments on digital platforms. Much of this difference comes from the difference in research methods and ways of understanding. The development of the theory of publicity stems from the compilation and generalization of past jurisprudence, an approach that is certainly valuable, but hundreds of years of history and numerous jurisprudence constitute a troubling quest for current rationality. In contrast, although the theory of natural monopoly has undergone its own development, the understanding of natural monopoly in terms of cost disadvantage has been a mainstream theory in economics since the 1960s and 1970s. Utilizing this well-established theoretical framework on a digital platform undoubtedly provides better certainty.

Indeed, the development of antitrust law itself is a good example. Antitrust law was created to regulate large-scale trusts such as railroads, steel and oil, but in the early days, antitrust law did not have a clear theoretical framework, and relied on contract and tort theories in determining restrictive competition, using whether it constituted a restriction on the freedom of contract and whether it infringed on the interests of competitors as the basis for judging the unlawfulness of the law, which resulted in disturbances in the application of the law. After the Harvard School introduced the "structure-conduct-performance" (SCP) paradigm in the 1950s and the Chicago School introduced the price theory in the 1970s, antitrust law has been deeply integrated with economics, and economic theories and tools have been utilized more often to conduct analysis of illegality. In particular, the consumer welfare standard developed by the U.S. antitrust laws based on the goal of economic efficiency has enabled modern antitrust law to obtain a coherent, feasible, unified, and objective analytical framework, which has become the theoretical foundation of modern antitrust analysis. However, if the antitrust law is summarized only from the perspective of jurisprudence, it is difficult to find consistency in the viewpoints of "protection of the competitive order," "protection of competitors," "protection of the competitive process," and so on.

2. Limited intervention. As mentioned above, the public nature theory emphasizes the public interest attributes of digital platforms, which brings all digital platforms into the scope of regulation and has the effect of intervening extensively in the market. In contrast, economic regulation restricts intervention to areas of cost disadvantage, which not only has a clearer scope of regulation, but also greatly reduces intervention in the market. Such limited intervention also has advantages:

First, limited intervention can better respond to the rapid innovation that characterizes digital platforms. Historically, digital platforms have been characterized by rapid innovation, with new technologies and business models constantly emerging. Today's social networks and algorithmic search engines have evolved from markets once known as "portals," such as AltaVista and American On Line. The once leading social networking sites, such as Friendster and MySpace, will soon be extinct. It should come as no surprise, then, that digital platforms will change in the near future. However, all ex ante regulatory measures are largely a static view of the industry. Because innovation is difficult to predict, targeted measures are difficult to presuppose. With extensive interventions, total government involvement in the industry leaves the industry inflexible enough to adapt to rapid change.

Economic regulation is likewise intervention, and there is a trade-off between the costs and benefits of regulation, i.e., the potential loss of short-term consumer welfare from not regulating versus the loss of, for example, the disincentive to innovation that regulation creates. The false cost analysis framework proposed by Judge Frank Eastwood for the enforcement of antitrust laws is equally applicable to the formulation of appropriate market regulation policies, i.e., measures that act on the market must take into account both false-positive and false-negative costs. In general, false negative costs are better neutralized by the market itself. At a more specific level, unregulated free competition has three corresponding potential advantages over economic regulatory measures: first, the likelihood that producers in an industry will have low marginal costs is higher than in the case of monopolies, because even if one firm cannot ensure low costs, its competitors may be able to do so. Two, the presence of competitors with similar costs reduces the information advantage of industry producers. Third, any direct, operational regulatory costs, such as the salaries of regulators and their staff, are avoided. Thus, greater preservation of the scope for market competition is certainly more desirable in markets where innovation is more rapid.

Second, limited intervention better reduces the costs associated with the rigidity of regulatory institutions and with regulatory capture. Once regulation begins, it has a tendency to become self-reinforcing. Regulators will continue to justify their existence, thus creating a "ratchet effect". Even in industries that were once natural monopolies (including telephones, airlines, railroads, and power generation), regulation is often reluctant to withdraw even when competition becomes feasible. This is because these organizations, and the people involved, depend on regulation for their very existence. Moreover, according to the classical theory of the economics of regulation, regulation creates monopolies and generates monopoly profits. Thus, regulation may be the result of active seeking by the regulated firms. When regulated firms can obtain the benefit of exemption from competition through regulation, they will do their best to prevent deregulation. Furthermore, consumers who are harmed are difficult to overcome because of the large number of people, the significant cost of negotiation and organization, and the difficulty in overcoming free-rider behavior, resulting in a lack of sufficient power to fight against the regulator and the regulated firms to effectively push for deregulation. Based on the combination of these aspects, when a slight or moderate level of intervention is implemented, it makes the likelihood of implementing a greater level of intervention in the future rise. Therefore, limiting the scope of regulation with a clearer theory can avoid or minimize these costs to a greater extent.

While regulation in response to market failures is not perfect, to a large extent it is a "necessary evil", the product of a trade-off between costs and benefits. The key, therefore, is not whether regulation is needed, but how to keep it within appropriate limits. The public nature of the theory of the existence of the definition of ambiguity and the broad scope of control, the combination of the two will make the control of the direction of over-intervention, resulting in the inhibition of market vitality of the consequences. On the other hand, the natural monopoly theory constructs the basis of control with the criterion of contestability and non-competitiveness, and the theoretical logic is more reasonable; it limits the direct control to a smaller scope, which can leave more space for free competition.

4.3 Realization of economic control on digital platform

Compared with the theory of publicity, economic control has better application prospects in digital platforms, and economic control itself has a reasonable theoretical framework, which provides a basis for constructing a complete control system for digital platforms.

4.3.1 Structural divestiture as a prerequisite for economic regulation

In understanding natural monopoly, it is necessary to distinguish between segments and industries with the attribute of natural monopoly, which is very crucial to understand the existing situation of digital platforms. While much of the literature often uses the term "natural monopoly" to refer to all activities in a given utility service supply chain, i.e., the entire industry of gas, electricity, water, etc., public policy in fact needs to separate activities in the supply chain that have the attributes of a natural monopoly from those that are competitive. That is, not all segments in a given industry may be natural monopolies. For example, in the electricity industry, generation is competitive and the electricity transmission network has natural monopoly properties. This understanding holds true for digital platforms as well. For example, even if Google's search engine has natural monopoly attributes, the products surrounding the search engine, such as Google Translate and Google Maps, are not ipso facto natural monopolies.

Structural divestiture is an important means of ensuring the separation of natural monopoly and competitive businesses of digital platforms, and a prerequisite for realizing economic regulation of digital platforms. Moreover, through structural divestiture, vertically integrated firms can be prevented from shifting profits, removing the inherent incentives to discriminate and realizing the non-discrimination obligations that are of particular concern to publicity theory. For example, Google has more than 70 separate products and services that are highly diverse and independent, but all exist within a single ecosystem. Within this ecosystem, Google retains all rights to access and generate data. It is inevitable, then, that discrimination arises when Google's own business conflicts with operators in the ecosystem. Only by divesting the products and services with natural monopoly attributes in it can the conflict of interest problem be fundamentally solved.

By divesting the natural monopoly business, it is also possible to address market concentration due to natural monopoly as well as the competition and societal problems derived therefrom, including preventing the expansion of the dominant position of the incumbent operator through cross-financing, protecting the resilience of the system, promoting diversity, and preventing the over-concentration of power, and so on. For example, the collapse of infrastructure services has a greater impact on economic activity when all operations are consolidated, and the more operations that are consolidated together, the greater the likelihood of system collapse. Structural divestitures can address these issues to some extent. Indeed, there has been an argument in the platform economy for unbundling platforms, in the hope that this would return the digital economy to its natural, legitimate and decentralized state, thereby fostering innovation and diversity of choice. However, this proposition leaves all platforms undifferentiated, depriving platform firms of the efficiencies that come with integration and thus resulting in excessive intervention in the market. Limiting structural divestiture to natural monopoly segments, on the other hand, provides a better balance between competition and efficiency.

In fact, there are many examples of structural divestitures to address competition concerns in traditional utilities, including the 1982 U.S. antitrust consent decree against AT&T, whose network was presumed to constitute a natural monopoly and was required to be broken up into multiple companies, each of which was responsible for a specific geographic area. There is no barrier to structural divestiture for digital platforms, and even though today's large platform companies have a very large number of business branches, often constituting a large ecosystem, they have evolved gradually from their initial core business. For businesses with natural monopoly attributes, because of their market position, there is no question of them being unable to survive even if they are divested. There have been attempts to do so in the digital economy. In the case of United States of America v. Microsoft Corporation (hereinafter referred to as the "Microsoft case"), which was once called the "antitrust case of the century", the court was prepared to split Microsoft into two parts, namely, application software and system software, to be operated separately. More recently, the U.S. Federal Trade Commission, in its antitrust complaint against Facebook, asked Facebook to divest its Instagram and WhatsApp businesses. Of course, these cases represent different spin-off ideas. the AT&T divestiture was a horizontal spin-off, with each of the spun-off companies retaining the incumbent local exchange carrier's monopoly on local services. The Microsoft and Facebook divestitures, on the other hand, were primarily based on business functions and were vertical spinoffs. For digital platforms, AT&T's divestiture may be more limiting because it is difficult for digital platforms to operate within a specific geographic area. Instead, a vertical spin-off based on the natural monopoly attributes of the different business functions themselves would be more appropriate.

4.3.2 Interoperability obligations at the heart of economic regulation

After the divestiture of natural monopoly businesses or segments has been realized, the problem of linking contestable and natural monopoly businesses arises, and the interoperability obligation is the basic rule to solve this problem. A natural monopoly means that the digital platform has a monopoly in the relevant market, and if it does not provide sharing of information and resources, the divestiture will not fundamentally solve the problem of competition in the market, and therefore it is necessary to ensure that the platform with a natural monopoly provides that infrastructure to new entrants on competitive terms. Through interoperability obligations, it is possible for the platforms in question to have access to foundational resources, and new market entrants will not have difficulty in developing because they are constrained by network effects and scale efficiencies. It is in this sense that the interoperability obligation becomes the core obligation of digital platform economic regulation. In the AT&T case, the U.S. court, while conducting structural divestiture, required not only that parts of the old Bell System be interconnected with each other, but also that they be interconnected with new competitors to provide switched access on an unbundled, fee-based basis. Thus, it can be argued that the interoperability obligation and structural divestiture complement each other and are the most critical mandatory rules for natural monopoly businesses.

Mandatory interoperability obligations are specific to natural monopoly platforms and do not constitute a universal obligation for all digital platforms. Competitive digital platforms have the autonomy to decide whether or not to adopt interoperability, depending on the competitive outcome that interoperability may bring. In general, interoperability can effectively integrate the networks of multiple platforms, thereby expanding network effects, and thus platforms have strong incentives to remain interoperable with each other. However, not providing interoperability can lead to differentiation advantages although it reduces the advantages of network effects. In the case of social media platforms, for example, whether they are doing short-video socialization, business socialization, or photo socialization, these products benefit from network externalities and often have no or limited interoperability with each other, but they are significantly different from each other, attracting different, albeit overlapping, user groups. It is therefore more appropriate to leave it to the market to decide freely whether to adopt interoperability, outside of natural monopolies.

The key to achieving interoperability is the establishment of technical standards. In the context of the Internet, interoperability is simpler than it used to be in the basic telecom industry and does not significantly increase costs. This is because interoperability in telecommunications requires not only the establishment of technical standards, but also a significant investment in costs such as dedicated lines and machines. In contrast, the marginal cost of providing interoperability in the Internet sector is likely to be zero or close to zero, and regulators require interconnections that are not too costly to regulate. At this point, third-party connectivity to digital platforms is primarily a requirement to exchange information with the platform, which sets the limits and policies for such communications through the application program interfaces (APIs) it provides to third parties, and is thus focused on what types of services, data, and network functionality that third-party application developers can send to and receive from the platform through the APIs. Thus, the main cost of digital platform interoperability is the establishment of open standards for the exchange of digital information to enable specific functions. Moreover, the functions that require the establishment of technical standards to be realized represent only a small part of the functionality of digital platforms. As for other functions of the platforms, they can still be designed in a way that the platforms consider to be beneficial to them, thus realizing differentiated advantages, such as better page layouts, stricter privacy protection, and so on. However, authorized platforms or natural monopoly platforms participating in interoperability must use the content covered by the standard technical transmission. As for the standard-setting organization for interoperable technologies, it can be composed of industry participants, consumer representatives, technical experts and government representatives, and its operation as well as expected effectiveness can be referred to the standard-essential patent-setting organization.

4.3.3 Building on Established Economic Control Measures

Divestment of the natural monopoly business of digital platforms and mandatory interoperability obligations are the most important aspects of realizing economic regulation of some digital platforms. However, economic regulation is a system with many more associated elements. For example, after divestment of natural monopolies, restrictions on market access are also needed to prevent inefficiencies arising from market entry. As another example, although interoperability is dominated by technical standards, it may also involve the price of the platform's products or services, and appropriate prices need to be set to incentivize the platform to reduce costs. In terms of market entry, while the natural monopoly attribute suggests that the market cannot be made more efficient through competition, qualifying for monopoly operations can be competitive, and thus there is the question of how to design an effective franchise contract. In this regard, there are further questions about whether the platform should be directly nationalized or privately operated, and the impact of considering different operating entities and governance models on operational efficiency. These are all issues that need to be addressed in the economic regulatory system of digital platforms, and it is impossible to expand on them all here.

Economic regulation is an important part of public policy in almost every country and region, and the practice of economic regulation of industries such as telecommunications, railroads, electricity, and water has been going on for many years before the development of digital platforms. Moreover, due to different understandings and practical considerations of the shifts in attitudes and the impacts of technological developments during this period, different countries and regions have had different institutional designs, leading to a large number of commonalities in addition to a diversity of experiences. For example, economic regulation usually entails setting price levels for products or services, and maximum price controls and their variants (e.g. income ceilings) are often used in actual policy. This method of pricing, known as incentive pricing, requires utilities to keep the price of essential services at no higher than a specified level for a specified period of time, thus providing an incentive to reduce costs in order to realize higher returns. These methods, which have been widely adopted by traditional industries, could also continue to be used in digital platform regulation. Overall, unlike the lack of systematicity and realism in the control measures advocated by the theory of publicity, the economic control of digital platforms can continue to be based on the economic control of traditional industries, and most of the aforementioned problems can be better solved as a result.

The construction of an economic control system is a systematic project that requires creative solutions to the challenges posed by digital platforms, which in itself is also the proper meaning of digital platform control policy in its formulation and development. Even in the economic control of traditional industries, there are also new issues arising from development in different industries that need to be fully considered in control measures. For example, one of the characteristics of electricity is that it is usually impossible to store it after it is produced, so it is necessary to consider the balance between the supply and demand of electricity to maintain continuity in the control of the power system. Digital platforms have their own characteristics and their economic regulatory systems face new issues. First, digital platforms are characterized by rapid innovation in technology and business models. Although the scope of economic regulation has been much smaller than that advocated by the theory of publicity, economic regulation, after all, replaces the market and has the weakness of rigidity inherent in regulation, which makes it difficult to effectively anticipate the development of the market. For example, when online taxi only appeared, it had a potentially natural monopoly tendency because of the application of Internet technology, which made it possible to operate directly on a very large scale, prompting the merger of the two largest online taxi companies of that year, Dripping and Express, in 2015. By September 2016, DDT had more than 15 million registered drivers nationwide. However, online car rental platforms have a bilateral market attribute, where the size of online cars and users has not only scale effects, but also cross-network externalities. It is also because of this attribute, when the aggregation platform through its own traffic portal will be small network car companies integrated together, it solves the user scale limitations faced by the network car, the network car itself scale limitations have also been broken. The business model innovation realizes the competitiveness of the net taxi platform. In this case, how to make the economic control system more flexible so that markets that no longer have natural monopoly characteristics can be deregulated in a timely manner requires solutions at the institutional level. Second, the regulation of specific business models is challenging, such as the free model that is widespread in the Internet sector. Price control is usually practiced in economic regulation, but when the price of a product or service is zero, such regulation becomes difficult to operationalize. Of course, the so-called "free" actually involves a bartering process. It is undeniable that when the price is zero, the focus of regulation can only shift to quality, but quality is not as intuitive and easy to quantify as price. Take search engines as an example, even if the total cost of the search engine plus a fair return on investment as the basis for pricing, and the price of search advertising services on the price growth of the implementation of price cap control, but the search engine for the search user is "zero price" to search users to determine the appropriate pricing of the search query or to regulate the quality of the search will be It would be difficult to determine appropriate pricing for search queries from search users or to regulate the quality of searches. These issues also need to be addressed in the implementation of economic regulation.