Digital Platforms Now Write Rules - Not Just Code

Digital platforms, such as Uber, Airbnb, YouTube, and Amazon, have become some of the largest and most influential companies in the global economy. Key to these platforms’ business models are network effects: the more participants that join and engage on the platform, the more valuable it becomes for all members.

To manage these interactions, platforms establish rules about access (who can join), behavior (what members can do), and the relationships between the platform, its members, and third parties. The goal of these rules is to overcome market frictions that may arise in platform interactions, such as high uncertainty, free-riding, or information asymmetry.

As private entities, digital platforms can adapt these rules at will. They can do so by adjusting their contractual agreements (e.g., “terms of use,” privacy policy, etc.) or their code (e.g., application programming interfaces or APIs, design features e.g., filters, etc.).

While these rules may be perceived as features of platforms’ business models, legal scholars view them as a source of regulation that is superimposed over and above regulation through law.

For instance, the “terms of use” policies of platforms like Facebook and YouTube outline what content members can create and share, which enforces limits on members’ rights as individuals to “allowable speech” and privacy.

Similarly, content and product recommendations on platforms like YouTube or Amazon are effectively rules that shape competition by benefiting the platform or some platform members at the expense of others.

I explore these questions both on a theoretical and empirical level. Drawing on insights from research on “rules as institutions” in political economy and law, I develop a theoretical framework to delineate different aspects of platforms’ rule-making.

For example, platforms can choose the type of rule to create, how to implement the rule, whom to enforce on, and how strictly. Building on this framework, I evaluate the impact of platforms’ rule-making choices on different stakeholders: individuals and firms offering services on platforms (known here as complementors), regulators, and social movement organizations.

I empirically examine the impact of how a platform decided to enforce rules (i.e., whom to target, and how strictly) on complementors and regulators across two studies. These studies employ recent advances in causal identification via a spatial difference-in-differences research design on very large-scale observational data — the entire population of more than a million Airbnb listings in Paris.

The studies are set in the context of the Paris short-term tourism rental accommodation market, and more specifically of properties’ compliance with the 2016 French ELAN law, which restricted short-term rentals of entire homes to 120 days annually unless licensed.

In this setting, the complementors are properties that offer short-term rental accommodation to tourists or “guests” on platforms like Airbnb, Booking.com, Vrbo/Abritel, etc. Notably, although this regulation applied citywide, Airbnb decided to only monitor compliance in the first four arrondissements (neighborhoods) of Paris from 1 January 2018.

From January 2019, the platform started sanctioning non-compliance by automatically blocking calendars of unlicensed properties that had reached the 120-day annual limit in the same four arrondissements. This system was finally extended to the rest of Paris from 1 July 2021.

The first study finds that after Airbnb voluntarily enforced the regulation, complementors’ regulatory compliance increased in the first four arrondissements. However, compliance was not perfect, as some complementors found ways to circumvent both the regulation and Airbnb’s voluntary enforcement by multihoming i.e., being listed on multiple platforms such as Vrbo/Abritel, apart from Airbnb.

The study indicates that complementors’ regulatory compliance matters to platforms because if there is too little overall compliance the platform itself may be at regulatory or reputational risk. Further, this research indicates that, in this instance, competition between platforms was detrimental to regulatory compliance, as complementors strategically utilized this competition, via multi-homing, to evade regulation.

The second study identifies an unintended consequence of Airbnb's enforcement choices: individual hosts outsourcing the management of their properties to professional property management firms, thereby creating a new class of complementors.

However, individual hosts that did not shift to these professional firms were worse off as they faced stiffer competition, within their arrondissements (neighborhoods), from these professionally managed properties.

Additionally, I theoretically model why some social movement organizations are more successful than others in mobilizing resources online (e.g., Black Lives Matter). Social movement organizations need information and connections to mobilize resources from individuals and organizations.

However, platforms’ rules govern how members interact and connect with other members on the platform. Therefore, contrary to extant research, the study finds that platforms’ rule-making choices not only enable but also constrain the ability of social movement organizations to mobilize resources because platform rules shape how their members share information, coordinate, and organize for collective action.

Platforms’ rule-making choices not only enable but also constrain the ability of social movement organizations to mobilize resources because platform rules shape how their members share information, coordinate, and organize for collective action.

In summary, platforms’ rule-making, while primarily driven to enable interactions on the platform to manage network effects (a market action) has non-market consequences by intermediating first, the regulatory compliance of complementors, and second, the information and communication environment of their members affecting members’ ability to organize for collective action.

I also show that platforms’ rule-making has implications for non-market actors such as regulators and social movement organizations. Beyond its theoretical implications for research, this thesis also contributes to the broader discussion on the role of platforms in society.