contents go

KDI - Korea Development Institute

KDI - Korea Development Institute

SITEMAP

HOT ISSUE

KDI FOCUS Competition Policy for Online Platforms’ Self-Preferencing Conducts August 21, 2024

KDI FOCUS

Competition Policy for Online Platforms’ Self-Preferencing Conducts

August 21, 2024
  • 프로필
    Min Jung Kim


Self-preferencing by online platforms encompasses a spectrum of conduct types, all with implications for both anti-competitive and pro-competitive effects. Therefore, it is desirable to maintain the current ex-post regulatory framework, which applies the rule of reason and intervenes only when such practices are deemed to be unjust or harmful to fair competition, rather than imposing a blanket ban. However, the timeliness and efficiency of enforcement mechanisms should be improved in light of the distinct characteristics of the online environments and self-preferencing conducts.


Ⅰ. Issue

As catalysts of innovation and competition, online platforms have enhanced consumer welfare by facilitating easier transactions, enabling instant information sharing, and introducing a wide range of new products and services. Deeply integrated into everyday life, the services provided by these platforms are diverse, covering shopping, internet search, social networking, messaging, content sharing, operating systems, app marketplaces, voice assistants, cloud computing, mapping, and advertising. As the dependence of economic agents on online platforms continues to grow, some platforms have rapidly expanded to enormous scales. As of 2023, half of the world’s top 10 companies by market value (Apple, Microsoft, Google, Amazon, Meta) are platform companies, commonly referred to as Big Tech (see Table 1).

Online platforms are increasingly central to daily life, with some experiencing exponential growth to reach enormous scales.

As a result, concerns are mounting over the concentration of economic power in a small number of dominant platforms. In parallel, there is increased attention to monopoly issues and competition policies related to online platforms, accompanied by active discussions on regulatory improvements. Over the past several years, competition authorities worldwide have initiated numerous investigations and lawsuits against online platforms for abusing market dominance. A particular focus has been on self-preferencing conducts, where these platforms favor their own products. A notable example is the Google Shopping case in the EU, which concluded in June 2017 after more than seven years of investigation into Google’s preferential treatment of its comparison shopping service in search results. Similarly, in June 2022, Bundeskartellamt, the German competition authority launched an investigation against Apple for allegedly favoring its apps by restricting user data tracking features. In South Korea, major lawsuits related to self-preferencing are ongoing, including the Naver Shopping and Video case (October 2020), and the Kakao Mobility case (February 2023).

As concerns grow about the over-concentration of economic power in a small number of online platforms, discussions on improving competition policies related to market dominance have intensified.

In addition, major countries have recently made legislative moves to strengthen regulations against monopolistic practices by online platforms, with new laws or bills specifically targeting self-preferencing conducts. Notable examples include the 10th Amendment to German Competition Act (GWB Digitalisation Act) and the EU’s Digital Markets Act, both targeting large digital platforms and effective from January 2021 and May 2023, respectively. In the UK, the Digital Markets, Competition and Consumers Bill passed through Parliament in May 2024, while in the US, the American Innovation and Choice Online Act was reintroduced in June 2023. These laws and bills adopt ex-ante designation to identify platform operators in advance for regulation and prohibit anti-competitive behaviors such as self-preferencing and limiting interoperability among designated business operators. In December 2023, the Korea Fair Trade Commission (KFTC) also announced plans to pursue the enactment of the “Platform Competition Promotion Act” (tentative title).

However, opinions are sharply divided on how to regulate self-preferencing, with views ranging from banning the sale of own products to prevent self-preferencing altogether, to considering self-preferencing as a legitimate aspect of the competitive process. The rationale behind revamping competition laws or introducing new regulations overseas is rooted in the concern that, while platform dominance has grown excessively large, current competition policies are inadequate to address the potential for rapidly escalating harms of monopolization. Conversely, some argue against further regulation, suggesting that because the platform market is sufficiently dynamic and continuously innovative, additional regulation could hinder incentives for innovation and harm competition in the long run. Against this backdrop, this paper proposes directions for improving competition policy to effectively address self-preferencing by online platforms.

As major countries strengthen regulations on online platform monopolies, self-preferencing practices have become a key area of focus.
This paper proposes directions for improving competition policy to effectively address self-preferencing practices by online platforms.


Ⅱ. The Dual Role of Platforms and the Concept of Self-Preferencing

Self-preferencing has been the subject of extensive discussion, but its concept remains loosely defined. In short, self-preferencing can be seen as a practice where a platform treats its own or affiliated companies’ products and services more favorably than those of competitors on its platform. In Korea, the ‘Guidelines for Review of Abuse of Market Dominance by Online Platform Operators,’ in effect since January 2023, also includes within the scope of self-preferencing the behavior of indirectly favoring oneself by preferentially treating the products and services of third parties with which one has a vested interest.

Self-preferencing arises from the dual role of platforms, an arrangement frequently compared to “a referee playing as a player.” Take, for example, an operating system platform that also provides apps or a marketplace platform that also acts as a retailer. At the outset, many online platforms offer free services to build a user base, but as they grow, they expand into adjacent sectors to generate revenue once they achieve a certain scale. As in Figure 1, self-preferencing occurs when platforms enter markets where they act as intermediaries and thus become competitors to their business users. In other words, platforms typically set rules—often unilaterally—while providing core services that mediate transactions for business users. When they simultaneously sell their own products and services on their platform, they become direct competitors to those users, potentially creating incentives to apply rules more favorably to themselves. However, platforms also have a strong incentive to avoid discriminating against business users, as they benefit from network effects from increased participation, creating conflicting incentives for platforms: more immediate financial gains from self-preferencing versus longer-term benefits from reputation management and ecosystem growth through fair treatment. As a platform secures sufficient network scale and attains a dominant market position, the incentive for fair treatment may

Without clear delineation, self-preferencing is a catch-all term for a broad spectrum of behaviors. Behaviors that prompt concerns about self-preferencing, including refusal to deal, margin squeeze, tying and bundling, and exclusive dealing, are also considered traditional anti-competitive practices. The problem is that treating it as a single category under competition law could lead to legal uncertainty. This is undesirable because, as an overly broad category, self-preferencing encompasses fundamentally different behaviors that should be subject to distinctive legal standards (Colomo, 2020). To ensure appropriate regulation, it is essential first to classify and specify the diverse conducts included under self-preferencing.

Self-preferencing refers to the conduct of a platform favoring its own or affiliated products and services over those of competitors, arising from the platform’s dual role.
Self-preferencing encompasses heterogeneous conduct types, so it needs to be categorized and specified for proper discipline.


Ⅲ. Categorization and Characteristics of Self-Preferencing

This paper categorizes self-preferencing into four types, by grouping similar conducts based on their patterns and nature: ① preferential placement through undisclosed algorithms, ② identifiable preferential placement, ③ discriminatory access to data, and ④ discriminatory access to other inputs and markets (Table 2). Types ① and ② involve preferential treatment regarding the placement of products or services in online spaces or on digital devices, while Types ③ and ④ pertain to access discrimination to inputs and markets.

In this paper, self-preferencing is categorized into four types.

Type ①: Preferential Placement through Undisclosed Algorithms 
Platforms are widely using algorithms to place and display intermediary products and services in limited online spaces—on the screens of digital devices shown to consumers. The first type involves behaviors related to placement through undisclosed algorithms, which can be further divided into (1) preferential treatment in the sorting and ranking of search results and (2) preferential treatment in other prominent placements, such as special or recommended products.

First, the most commonly observed type of self-preferencing involves platforms setting algorithms to give their own products and services higher search rankings. A representative example is the Google Search (Shopping) case in the EU, where Google was found to have systematically favored its comparison shopping service and lowered the rankings of competitors’ services in general search results using algorithms, thereby abusing its market dominance as a search engine. Second, platforms favoring their in-house products for prominent positioning present another concern. A notable enforcement case is the EU’s Amazon Buy Box case, which concluded with a commitment decision in December 2022—where the case was resolved by the platform proposing reasonable voluntary corrective measures. The investigation found that Amazon unfairly prioritized its in-house retail business or sellers utilizing its logistics and delivery service, Fulfillment by Amazon (FBA), in the selection criteria for Buy Box and Prime listings, offering significant visibility on product pages.

First, preferential placement through undisclosed algorithms involves favoring a platform’s own products in search rankings or other prominent positioning.

Type ②: Identifiable Preferential Placement
The second type also involves platforms favoring themselves in placing intermediary products and services. Unlike Type ①, this type can be immediately observed externally and typically occurs when platforms that provide digital devices or operating systems give preferential placement to native software. Type ② can be divided into two categories: (1) pre-installation, where the platform pre-loaded its first-party complementary services on digital devices such as PCs, smartphones, or smart speakers; and (2) default settings, where the platform’s services are pre-configured to run or connect automatically on programs or apps like browsers or email services.

Regarding pre-installation and default settings, issues have prominently arisen concerning proprietary apps (US House Subcommittee on Antitrust Report, 2020). On iPhones, Apple pre-installed only its first-party apps like Apple Music, setting them as the default service accessed via its voice assistant, Siri. Also, Apple’s web browser and email apps were set as defaults, and until the release of iOS 14 in September 2020, it was impossible to switch to third-party apps.

Second, identifiable preferential placement involves platforms pre-installing or setting their own software as default.

Type ③: Discriminatory Access to Data
The third type pertains to platforms leveraging data inaccessible to third-party business users to favor their own products and services. Platforms have access to a vast amount of information gathered from transactions carried out by all business users on their platforms. They are incentivized to use this wealth of information for their goods and services in pricing, inventory management, quality improvement, or launching new products. A frequently raised issue is platforms using sensitive business information from third-party users tolaunch imitation products that compete with those users. This type differs from other behaviors as it is about platforms’ exclusive use of their superior access to non-public data rather than blocking thirdparty access to data they might otherwise obtain. From the outset, business users are barred from accessing each other’s non-public data under privacy and data protection laws. Therefore, to ensure equal treatment between first-party and third-party products on a platform, the only viable option would be to level the playing field by prohibiting platforms from using confidential data, just as third parties are restricted.

A key enforcement case on data access discrimination is the EU’s Amazon Marketplace case. Conducted concurrently with the Amazon Buy Box case, this investigation also concluded with a commitment decision. The European Commission found that Amazon had access to extensive datasets related to the activities of third-party sellers, including order and shipment volumes, seller revenues, and shipping data. The authority determined that Amazon systematically leveraged this information to favor its own retail business operations in direct competition with those third-party sellers on its platform.

Third, discriminatory access to data occurs when a platform exclusively uses non-public data to develop its own products.

Type ④: Discriminatory Access to Other Inputs and Markets
The fourth type involves platforms providing their own products and services with preferential access to other inputs (except data) or to the market itself. When business users offer complementary products and services to the platform (typically providers of digital devices or operating systems), (1) discriminatory behavior may occur in access to hardware and software functionalities. That is, platforms may grant access to specific functions earlier or exclusively or offer superior functionalities to their own products. More generally, (2) platforms may discriminate in granting access to various inputs, such as advertising and marketing tools, delivery services, and after-sales services. Lastly, (3) discrimination may occur in access to the market itself, where platforms allow only own products and services to be sold.

Also common is discriminatory behavior in access to hardware and software functionalities, as exemplified by the ongoing investigation of Apple’s App Tracking Transparency Policy in Germany. Introduced in April 2021, this policy requires third-party apps to obtain user consent for tracking but exempts Apple’s first-party apps. For discriminatory access to delivery services, Amazon provides a notable example. During the COVID-19 pandemic, Amazon temporarily suspended deliveries of non-essential goods due to a sudden surge in sales but continued delivering similar non-essential items, such as hammocks and fish tanks, sold by its retail divisions. Lastly, concerning market access discrimination, Apple arbitrarily enforced app marketplace rules to favor its native apps or those with vested interests.

Fourth, discriminatory access to other inputs and markets arises when a platform grants preferential access to inputs and market for its own products.


Ⅳ. The Economic Effects of Self-Preferencing

Anti-competitive Effects
A major concern for self-preferencing arises when platforms leverage their dominance in the intermediary service market—their core business area—to unfairly gain an edge or exclude competitors in adjacent markets (Figure 1). This practice uses dominance in one market as leverage to gain a comparative advantage in another, even if it does not establish dominance there (Lee, 2020 and others). Through self-preferencing, platforms can effectively raise rivals’ costs or foreclose competitors.

The initial impact of a platform’s self-preferencing on competing business users varies depending on the type of behavior (Table 3). First, preferential placement (Types ① and ②) primarily reduces the business opportunities for competitors. When competitors’ products or services are displayed in relatively unfavorable positions, their sales prospects decrease even under the same competitive conditions (price, quality, etc.). Even if competitors are not entirely pushed out of the market, they may face increased actual competition costs to reach consumers, such as higher advertising expenses. Next, discriminatory access to data (Type ③) can initially weaken the innovation incentives of competing business users. The reasoning is that when a platform launches own competing products, it erodes business users’ profits, diminishing their incentive to participate in the platform and innovate. Lastly, discriminatory access to other inputs and markets (Type ④) resembles refusal to deal or supply, mainly leading to reduced quality or business opportunities for competitors. Competitors may experience lower quality if they cannot access hardware and software functionalities or lose sales chances if they cannot access other platform services or the market itself.

The primary impact of self-preferencing conduct on competing business users varies depending on the type, potentially reducing business opportunities, weakening innovation incentives, or lowering quality.

The initial effects outlined above may ultimately lead to the hindrance of price, quality, and innovation competition in adjacent markets. As competition in adjacent markets is restricted, prices may increase, and product diversity may decrease.11) Additionally, competitors might experience delays in decision-making or, in the long run, face reduced incentives for innovation and market entry.

Moreover, preferential placement (Types ① and ②) can mislead consumers and impede their rational decision-making, as platforms present information in a way that favors their interests. This form of exploitative abuse (exploitation of consumer surplus) can lead consumers to make choices that do not necessarily align with their best interests. By presenting certain products more prominently, consumers may perceive these products as being of higher quality or better aligned with their preferences. Also, they may exhibit a behavioral bias towards platform’s own products that are pre-installed or set as defaults.

As a result, price, quality, and innovation competition in adjacent markets may be stifled.
Preferential placement also has the characteristics of exploitative abuse.

Pro-Competitive Effects
On the other hand, self-preferencing by platforms can also yield positive effects, including lowering product prices, maintaining or improving quality, reducing consumer search costs, increasing product diversity, and promoting competition and innovation (Table 4). Specifically, when platforms offer own products, they may reduce prices by eliminating double marginalization or enhance the quality of intermediary services (including inputs) and their proprietary products through more efficient coordination of decision-making. Additionally, increased profitability from self-preferencing can incentivize platforms to invest further in intermediary services, potentially enhancing their quality.

For Types ① and ②, self-preferencing can increase consumer welfare when preferential placement boosts the demand for platform’s own products or those of affiliated third parties, leading consumers to purchase lower-priced or higher-quality items. However, such efficiency gains may come with the trade-off of limiting consumer choice by steering them away from their preferred products. Particularly in Type ①, the efficiency gains from self-preferencing may be more limited than other types. If the platform’s own products are genuinely superior in price and quality, they would appear at the top of search results even without self-preferencing through algorithm rules. In addition, Type ② may reduce consumer search costs. Preinstallation and default settings, like tying or bundling, provide a onestop shopping channel that reduces the costs for consumers to find the desired products.

Regarding Type ② and part of Type ④ (discriminatory access to functional inputs and the market itself), if compatibility, uniformity, and seamless functionality between related products are important, self-preferencing can help mitigate quality control issues. Especially if the quality of complementary products and services is crucial, but those offered by third parties are substandard or difficult to verify, self-preferencing is more likely to enhance efficiency. This can help not only in improving the overall quality of products and services but also in protecting security and privacy.

Lastly, Type ③ generates efficiency effects comparable to selling own products. If a platform effectively utilizes the information collected through operating intermediary services to identify business opportunities in niche markets and launches differentiated products from third-party business users, it could increase product diversity. If the platform’s introduction of those own products puts competitive pressure in that market, it can promote competition or innovation by offering better value for money and increasing transaction volumes. Moreover, if innovative own products grab consumer attention, this could trigger spillover effects that expand overall market demand, creating incentives for competitors to innovate.

Conversely, self-preferencing can also have positive effects, such as lowering product prices, maintaining or improving quality, reducing consumer search costs, increasing product diversity, and promoting competition and innovation.

Overall Effects
In summary, self-preferencing practices by platforms can yield both anti-competitive and pro-competitive effects. Self-preferencing does not necessarily lead to anti-competitive effects, and their overall impact cannot be conclusively determined. For instance, different countries have reached varied conclusions on Google’s preferential treatment of its own services in general search results. The EU found that Google’s self-preferencing excluded competitors and undermined fair competition in the comparison shopping service market. However, in the US (following the Federal Trade Commission investigation concluded in 2013) and the UK (High Court ruling in 2016), Google’s actions—in the case of the UK, preferential ranking of Google Maps in general search results—were deemed unlikely to result in anti-competitive foreclosure. Even where some adverse effects were observed, they were justified by efficiency gains in enhancing the quality or convenience of general search services (Jacobson and Wang, 2023).

In conclusion, the economic effects of self-preferencing should be evaluated through a comprehensive analysis that considers the specific competitive dynamics in the adjacent market where self-preferencing occurs, the characteristics of the relevant products and services, and the features of the platform services and market. The likely outcomes of self-preferencing suggest that when a platform’s own products are less favored or of inferior quality compared to third-party products, the anti-competitive effects tend to be more pronounced.16) Conversely, when the competition in the adjacent market mediated by the platform is limited, the efficiency gains from self-preferencing may be more substantial.

Self-preferencing can have both anti-competitive and pro-competitive effects, and the overall impact cannot be definitively determined.


Ⅴ. Competition Policy Direction

Basic Regulatory Direction for Self-Preferencing
The online platform market poses a heightened risk of monopoly power abuse. This risk arises from the market’s inherent tendency toward monopolization, driven by its winner-takes-all nature and the platforms’ roles as rule-setters and gatekeepers. In addition, platforms’ extensive capabilities of data collection and use and consumer behavioral biases may further exacerbate this risk. Regarding self-preferencing, the negative impacts of preferential placement, for example, may be amplified by consumer passivity in exposure to product placements and behavioral biases.17) Therefore, self-preferencing by online platforms requires appropriate regulatory oversight.

However, the downright prohibition of selling own products or self-preferencing is not advisable, as these practices can also enhance efficiency. Continuous technological advancements and innovation characterize the online platform market, driving the development of new business models and increasingly blurring market boundaries, which makes it difficult for the incumbent platforms to maintain a solid user base. Preemptive prohibition of self-preferencing without fully considering such dynamic features could inadvertently stifle innovation. While insufficient regulation against self-preferencing is problematic, the possible adverse effects of excessive regulation should not be overlooked. Competition policy must strike a delicate balance between mitigating the harms of monopolization and preserving incentives for innovation and healthy competition on platforms, which requires prudence to discern harmful selfpreferencing behavior from benign ones.

Self-preferencing by online platforms should be regulated appropriately.

Thus, it seems appropriate to regulate self-preferencing by applying the rule of reason, similar to the existing regulation of abuses of market dominance. Since pro-competitive effects may prevail depending on market competition conditions, platform characteristics, the nature of goods or services, and specific methods of self-preferencing, a case-by-case evaluation of economic effects is essential. Incentives other than excluding competitors may drive self-preferencing, and its presence alone does not automatically translate to significant anticompetitive effects. Even when anti-competitive effects are present, they must be weighed against potential positive efficiency gains. In conclusion, no substantial changes to the current ex-post regulatory approach under competition law appear necessary to address selfpreferencing by online platforms.

Nevertheless, it calls for careful attention to the current and potential obstacles in enforcing competition law. For identifiable forms of preferential placement (Type ②) or discriminatory access to functional inputs and markets (part of Type ④), platforms often justify their actions on technical grounds, such as security maintenance or privacy protection, which poses challenges to proving anti-competitive intent or impact. Moreover, in cases of preferential placement through undisclosed algorithms (Type ①) or discriminatory access to data (Type ③), the use of unobservable algorithms or proprietary data can complicate not only the proof of anti-competitive effects but even the determination of whether such actions have occurred. Given these difficulties, there are concerns that it may be difficult to intervene and respond promptly to the rapidly spreading harms of monopolization under current laws. The next section will discuss measures to enhance the timeliness and efficiency of enforcement.

As overregulation may result in even greater unintended side effects, it would be desirable to apply the rule of reason to self-preferencing, regulating it only when deemed unjust.While the current ex-post approach for competition law enforcement does not require material changes, it is necessary to enhance its timeliness and efficiency.

Enhancing the Timeliness and Effectiveness of Enforcement
First, the current framework for regulating the abuse of market dominance offers the advantage of flexibility, allowing for responses tailored to the specifics of each case. However, timely intervention remains problematic, as defining markets and determining market-dominant operators are time-consuming. Rather than insisting on excessive preciseness in market definition and dominant operator identification as prerequisites for assessing anti-competitive behavior, the policy focus should shift toward evaluating the economic effects of such behavior. This approach is more versatile, enabling simultaneous assessment of competition relationships and anti-competitive effects instead of rigidity requiring a complete analysis of competitive dynamics before moving to the assessment stage. In addition, it may be necessary to revise the criteria for assuming and recognizing market dominance. Enhancing the accessibility of market data for competition authorities by increasing budgets and specialized personnel for data-related tasks would also facilitate more timely assessments of market and dominant players.

As for the regulatory scope, there has been active discussion about designating business operators in advance, similar to practices observed internationally. While this approach is beneficial in timeliness and prevention, it also carries risks of suppressing innovation and healthy competition. If adopted, it would be prudent to restrict the scope to those with limited effectiveness and greater enforcement difficulties, rather than blanket application. For instance, preferential placement of own products using undisclosed algorithms (Type ①) or simple imitation of competitor products using proprietary data (part of Type ③) are likely to yield limited pro-competitive effects. Also, as these behaviors represent new types of conduct distinctive from traditional abuses of market dominance, their enforcement presents particular challenges. Even with the ex-ante designation, thorough evaluations of their economic effects should be conducted on a caseby- case basis.

Undue strictness should be avoided in market definition and assessments of market dominance.
For the implementation of ex-ante designation, it would be advisable to limit its scope to certain conduct types where efficiency effects are limited and enforcement is difficult.

Second, if self-preferencing is stipulated as abuse with potential anticompetitive effects, it is essential to clearly and precisely delineate the various forms of such behaviors, considering their heterogeneity. As previously noted, treating self-preferencing as a single legal category could hinder effective enforcement. Therefore, there is a need to refine the broad definition in the “Guidelines for Review of Abuse of Market Dominance by Online Platform Operators” to reflect specific cases that have emerged so far, similar to the categorization in this paper. Even with additional legislation or ex-ante designation for stronger enforcement, prohibited conducts should be specified in detail.

Since competition authorities are expected to face challenges even in the recognition and proof of self-preferencing behaviors, not just in evaluating their anti-competitive effects, efforts to mitigate these enforcement difficulties are necessary. Particularly for Types ① and ③, where information asymmetry is significant, effective measures should be put in place to allow competition authorities or thirdparty oversight bodies to swiftly access platform algorithms or data when suspected of problematic behavior. Take the example of the EU’s Digital Markets Act. It boosts the effectiveness of enforcement by allowing competition authorities to request access to all relevant data, algorithms, and testing information from companies. This Act also empowers authorities to impose substantial fines (up to 1% of global annual turnover) or penalty payments (up to 5% of global average daily turnover) for non-compliance. Another worthwhile consideration is obliging platforms to retain data, algorithms, and related information for a specified period.

Explicitly banning selfpreferencing requires a clear description of the specific conducts involved.
To mitigate enforcement challenges, it is worth considering granting competition authorities timely access to platform algorithms and data or imposing obligations on platforms to retain relevant information.

Easing the burden of proof on competition authorities is one viable option. Some have proposed requiring firms to prove that their selfpreferencing behavior does not have anti-competitive effects or that the efficiency benefits outweigh any anti-competitive impacts. For example, the 10th Amendment to German Competition Act requires firms to provide objective justifications for prohibited conducts. However, if the burden of proof is shifted to business operators, this requirement should be limited to types that are expected to cause significant anti-competitive effects and are considered difficult to regulate under existing methods, ensuring that this approach does not effectively operate as a per se illegal.

Lastly, commitment decisions could also prove beneficial. As demonstrated in the EU’s Amazon Buy Box and Marketplace cases, commitment decisions allow online platforms to propose voluntary remedial measures, facilitating the swift resolution of anti-competitive self-preferencing practices. Provided thorough assessment of their effectiveness, commitment decisions can serve as a valuable supplementary tool in enforcing competition law, which can often be a lengthy process.

If the burden of proof is to be shifted to businesses to ease the load on competition authorities, it should be limited to certain types.
A commitment decision can be an effective tool to complement the enforcement of competition law.


CONTENTS
  • Ⅰ. Issue

    Ⅱ. The Dual Role of Platforms and the Concept of Self-Preferencing

    Ⅲ. Categorization and Characteristics of Self-Preferencing

    Ⅳ. The Economic Effects of Self-Preferencing

    Ⅴ. Competition Policy Direction
related materials ( 9 )
  • Key related materials
Join our Newsletter

World's Leading Think Tank, Korea Development Institute

Security code

We reject unauthorized collection of email addresses posted on our website by using email address collecting programs or other technical devices. To access the email address, please type in the characters exactly as they appear in the box below.

captcha
KDI Staff Information

Please enter the security code to prevent unauthorized information collection.

KDI Staff Information

Please check the contact information.

OK
KDI Staff Information

Please check the contact information.

OK