Taiwan Fair Trade Commission’s Penalty on Digital Platform — the Foodpanda Case (Taiwan)

Aaron Chen & Sally Yang

The Fair Trade Commission (TFTC) held in its September 23, 2021, Gong-Chu-Zi 110066 that Foodpanda Taiwan Co., Ltd.’s (hereinafter referred to as Foodpanda) restrictions on its partnering restaurants to price uniformly in-store and on platform, and to mandate the allowance of self-pick-up-order from customers constitute an improper restriction on trading counterparts’ business activity as part of the conditions for transaction. This causes restrictive competition and thus be subject to a fine of NT$2 million for violating Paragraph 5, Article 20 of the Fair Trade Act.

TFTC’s Reasoning

1. Market definition and market power: the TFTC held that since delivery platforms are essentially transactional platforms due to their bilateral nature, such platforms constitute a market themselves. Accordingly, the market definition, in this case, is “food delivery platforms,” whose geographic market is worldwide. Furthermore, it was determined that since Foodpanda has a high market share and a massive number of customers, which include consumers and restaurants, the competitive advantage gained from the network effect indicates its considerable market power.

2. Determining Vertical Restraints and Restrictive Competition

A. Regarding Foodpanda’s “restriction on its partnering restaurants to price uniformly in-store and on platform,” the TFTC considers it a “most-favored-nation clause,” which lacks the reasonable cause of promoting competition and will produce the following anticompetitive effects:

(1) The restaurant cannot set different prices for customers using Foodpanda’s platform and those who don’t. Thus, these restaurants can’t implement selective marketing promotion per their own strategies, which restricts competition between direct and platform sales channels;

(2) Since restaurants using Foodpanda’s platform have to pay the platform’s commission, the burden of platform costs, due to the abovementioned restriction, will shift evenly towards those using and not using the platform.

(3) Since the platform’s commission are evenly spread throughout restaurants’ consumers, the commission charged does not shift consumers from using food delivery to dining in-store or use pick-up order. As a result, the platform has no worry when adjusting its commission. Thus, it has high incentives and the ability to raise its commission, giving it strong pricing capabilities.

(4) It indirectly guarantees that the restaurants’ price won’t be lower in other food delivery platforms than that of Foodpanda’s, which is effectively similar to a “cross-platform price parity agreement.” This weakens other delivery platforms’ incentives and ability to compete with Foodpanda.

B. Regarding Foodpanda’s mandate “not to refuse self-pick-up-order from customers,” the TFTC considers it forcing restaurants into self-competition, whereby commissions are charged for Foodpanda’s “self-pick-up-order” despite its ineffectiveness in reaching new customers — the “self-pick-up-order” function is used mostly by existing customers. Moreover, the platform encourages its customers to use the “self-pick-up” function (e.g., 21% discount), resulting in more restaurants customers opting for such an option. Thus, it lacks the reasonable cause of promoting competition and effectively restricts competition.

Noteworthy aspects of this case

1. This case is among the few rare cases TFTC punishing digital platforms. The disposition of this case has detailed the economic scale of platforms, characteristics of the network effect, and the market definition of bilateral platforms. Thus, the case should be of considerable value regarding TFTC’s future enforcement in the field of the digital economy.

2. The TFTC noted in its September 17, 2021, press release that the country’s two major food delivery platforms — Foodpanda and UberEats — are not violating the Fair Trade Act by offering a discounted commission for restaurants with whom they have successfully reached exclusive dealing agreement As such, an exclusive dealing in and of itself is not per se illegal. However, the TFTC held that, as the number of exclusive dealings increases, or if other substantive and mandatory exclusive-transactions are made, there remains a likelihood of violation. If so, the TFTC will re-examine and intervene in due course.

3. Except for explanations on the characteristics of digital platforms and the potential impact on competition, TFTC’s competitive analysis on the vertical restraints, in this case, is not essentially different from past analyses of the same type. Such analysis uses a method similar to the “rule of reason,” which takes into consideration the totality of factors such as the potential impact on competition, whether a reasonable cause of promoting competition exists, and several other factors. Thus, businesses carrying out vertical trading restrictions should consider their purpose, reasoning, and effects to minimize risks.

Issues worthy of further discussion

1. The TFTC has rarely hidden the market share of the penalized business since market share is an investigative result, not a business secret. In this case, however, the TFTC has veiled Foodpanda’s and other platforms’ market share, which is a pity as no reference can be made by other businesses on how digital platforms’ market share were determined and how large a market share of a non-monopoly digital platform constitutes “businesses with considerable market power. ”

2. Vertical restraints should generally be analyzed primarily through the effects of inter-brand Still, the TFTC seems to focus excessively its analysis on vertical restraints per se and rarely addresses inter-brand competitions.

Vertical restraints inevitably restrict the counterparty’s business or competitive freedom. However, such restrictions are not considered by the Fair Trade Act to be per se illegal; further competitive analysis is required in finding whether anti-competitive concerns exist. As such, analysis of vertical restraints should focus first on the effects of “inter-brand competitions,” and only after that should intra-brand competitions be taken into consideration[1].

The TFTC held that the market definition of this case is “food delivery platforms,” thus, an analysis of “inter-brand competitions” should focus on the effect of Foodpanda’s restrictions of its partnering restaurants on Uber Eats or other “food delivery platforms.” Although the TFTC has held that Foodpanda’s “restriction on its partnering restaurants to price uniformly in-store and on platform” indirectly guarantees that partnering restaurants’ pricing won’t be lower in other food delivery services than that of Foodpanda’s — which weakens other delivery platforms’ incentives and ability to compete with Foodpanda — it should be easy for such an analysis to be corroborated with substantive evidence. For example, one can investigate whether restaurants on Foodpanda and Uber Eats are priced uniformly or whether Uber Eats’ ability to compete via methods other than pricing is damaged due to Foodpanda’s restrictions. However, it is unclear from the disposition whether such evidence was even gathered or used.

Furthermore, if restaurants feel that the aforementioned restrictions could negatively affect their profitability, unless there are obstacles preventing conversion, they can easily switch to another platform like Uber Eats. If restaurants can easily ‘hop’ to another platform, is it the case that Foodpanda’s restrictions damage inter-brand competition? Would it not, on the contrary, promote inter-platform competition? There seems to be room for discussion.

From the perspective of bilateral platforms’ characteristics, the network effect stems from the fact that value for users on one end of the platform increases in relation to the size of users on the opposite end. For example, having more restaurants meant attracting more consumers, and vice versa. Therefore, the number of users on one end of the platform may affect the number of users on the other, which ultimately affects the platform’s competitiveness. Since users’ incentives on each end of the platform differ, delivery platforms compete by opting for different strategies to attract users on each end. Unless Foodpanda is a monopoly and so isn’t worried about losing customers, the reason Foodpanda created its restrictions — knowing that restaurants may switch platforms — is its strategic decision to lower its price to attract users on the consumer end. On the other hand, Uber Eats opts for an opposite strategy whereby the focus is gaining more users on the restaurant end of the platform in hopes that a diversified list of restaurants can attract users on the consumer end. In simpler terms, due to the differing strategic decision between the two major food delivery platforms, as long as there are no difficulties switching platforms, Foodpanda’s restrictions may ultimately increase competition between food delivery platforms.

3. Alternatively, TFTC’s penalization against Foodpanda may be counterproductive in that it restricts available strategic decisions for these platforms, which ultimately leads to price uniformity. For example, suppose consumers can no longer freely select platforms based on restaurant prices and other service qualities due to price uniformity after TFTC’s disposition. Would that not damage consumers’ choices? If so, it is worth contemplating whether TFTC’s involvement in the food delivery platform’s market competition was premature. Moreover, it is worth considering whether it is appropriate to focus on the unfair effects of only one end of the platform’s users.

(The authors’ opinions do not represent the position of this law firm.)

[1] See Min-Jye Huang, The Attitude of Taiwan’s Courts toward Vertical Trading Restrictions and the Development Trend of Their Judgments, 29(2) Fair Trade Quarterly 44 (2021); Andy C. M. Chen & Yi- Yu Lin, Determine the Legality of Non-price Vertical Restraints under the Rule-of-reason Standard 9, Fair Trade Commission research project (2013); Tzu-Shun Hu, ECONOMIC ANALYSIS OF COMPETITION LAW 608 (2019).