We study data portability in a model of dynamic price competition where product quality improves for each customer as personal data accumulates with usage. This can reflect AI-driven personalization based on customer data or customers embedding their own data into the firm's product through usage. We show that when consumers are forward looking and firms can personalize their prices, each firm does best by offering data portability so that its customers can freely take their data to a rival. By contrast, when consumers are myopic, firms no longer always benefit by offering portability, and there are equilibria where neither firm offers data portability. A data portability mandate is therefore redundant when consumers are forward looking, but becomes potentially relevant when consumers are myopic. However, with myopic consumers, a data portability mandate can also cause a complete market breakdown if consumers' initial willingness to pay for firms' products is below cost. We extend these results to a setting without personalized prices.

How does a firm decide whether to employ professionals and control how they deliver services to clients, or to operate as a platform enabling independent professionals to provide services directly to clients? Similarly, how does a manufacturer decide whether to allow sales agents to choose certain costly actions (e.g. kickbacks to clients) or to take control of these actions itself? We answer this question using a principal-agent framework in which both the principal and the agent must be incentivized to carry out investments (or effort) that increase the revenue they jointly create. Our theory explains when the principal should take control over a particular decision ("control") or should instead allow the agent to make the decision ("enable"). It does so both for the case when there are multiple such transferable decisions for a single agent, and for the case when there are many agents and one transferable decision for each. We also consider the possibility of cost asymmetries between the principal and the agent, spillovers across agents, and the misclassification of the principal as an employer even though agents are allocated the relevant control rights. Finally, we explain how the "control vs. enable" choice and its associated tradeoffs differ from the classic "make vs. buy" choice.

We produce updated rankings of economics journals based on established and new methodologies, and use these rankings to document the spectacular rise of the new society journals in economics. We show that while several factors (editor reputations, editor experience, citations from parent journals, and the number of articles published) help determine these journals’ impact factors, none help explain why the new journals outperform natural comparison journals. However, soliciting top authors connected to the editors can explain their outperformance. We also consider factors such as fast turnaround times, the transfer of referee reports, and associations leveraging their reputations.

Rational agents can be induced to join a platform despite being better off without it, when each agent's outside option worsens as other agents participate on the platform. Such platform traps apply to sellers on dominant marketplaces and users on social networks, among other settings. We provide a dynamic theory of a platform trap in which the platform exploits a form of collective action problem through one (or more) of the following features: (i) the ability to make dynamic price adjustments; (ii) the combination of on-platform and off-platform negative externalities; (iii) favorable equilibrium selection when platforms make private offers.

This paper examines competition policy implications of the rapidly expanding Artificial Intelligence (AI) sector. We analyze the vertical AI technology stack and data feedback loops to address three key questions: the potential for market concentration in core AI services, AI's likely impact on existing market structures, and emerging competition policy challenges. We identify key risks to competition in the AI sector, ways in which AI may disrupt some existing platforms, how AI could lead to new types of gatekeepers, and some novel competition policy concerns raised by AI.

Choosing how easy to make it for buyers to discover new sellers is a key design decision for platforms. On one hand, enabling more discoverability generates more transactions and can be more attractive for sellers because they anticipate being discovered by new buyers. On the other hand, discoverability can make sellers more reluctant to participate because they anticipate their existing buyers will discover and purchase from other sellers. We model this fundamental trade-off and study how a platform’s optimal level of discoverability depends on various factors: the degree of substitutability between the sellers’ products; the standalone value of the platform’s tools; the size of the platform’s installed base of buyers; the nature of the platform’s fees, including its ability to charge differential fees depending on whether a transaction is with a seller’s existing buyers or not; the number of sellers; and asymmetries between sellers’ sizes. This paper was accepted by Joshua Gans, business strategy. Supplemental Material: The online appendix is available at https://doi.org/10.1287/mnsc.2023.02362 .

A key issue for the design of online marketplaces is addressing leakage. Buyers may use the marketplace to discover a seller or to obtain certain conveniences, but the seller may then want to take transactions off the marketplace to avoid transaction fees. Assuming buyers are heterogenous in their switching cost or inconvenience cost of purchasing directly, we provide a model in which there is partial leakage in equilibrium. We use the model to analyze the trade-offs associated with different strategies the marketplace can use to attenuate the effects of leakage: investing in transaction benefits, limiting communication, charging referral fees, using price-parity clauses, introducing seller competition on the marketplace, and hiding sellers that try to induce too much leakage. This paper was accepted by Joshua Gans, business strategy. Supplemental Material: The online appendix is available at https://doi.org/10.1287/mnsc.2023.4757 .

We provide a general framework to analyze competition between oligopolistic platforms in competitive bottleneck settings, allowing for various pricing and non-price design choices by platforms, and a range of buyer-seller microfoundations. We show that the equilibrium choices by platforms are distorted against sellers' interests in a way that is harmful to welfare (e.g., setting excessive commission fees), and that platform entry can exacerbate this distortion. We also characterize how buyer-side heterogeneity in interaction benefits, partial market coverage, platform asymmetry, and cross-platform spillovers further amplify or mitigate welfare distortions. Based on our findings, we discuss policy implications for mobile app platforms.

We consider platforms that help consumers discover and transact with suppliers. Such platforms have come to dominate many sectors of the economy, raising issues about the high fees they charge suppliers, especially since they tend to commoditize the suppliers they aggregate. We show that in a baseline setting, the efficient platform fee is determined by a simple formula: it equals the platform's marginal cost plus the difference between suppliers' markups on the direct channel and suppliers' markups on the platform. We explore the extent to which this simple formula provides a robust cap for regulating the platform's fee more generally.

We provide a general framework to analyze competition between any number of symmetric two-sided transaction platforms, in which buyers and sellers can multihome. We show how key primitives such as the number of platforms, the fraction of buyers that find multihoming costly, the value of transactions, and the degree of user heterogeneity jointly determine the level and structure of platform fees. Even though platform entry always reduces the total fee level, whether it shifts the fee structure in favor of buyers or sellers depends on whether most of the buyers are singlehoming or multihoming. (JEL D43, L11, L13, L40)

We model dynamic competition between firms which improve their products through learning from customer data, either by pooling different customers' data (across-user learning) or by learning from repeated usage of the same customers (within-user learning). We show how a firm's competitive advantage is affected by the shape of firms' learning functions, asymmetries between their learning functions, the extent of data accumulation, and customer beliefs. We also explore how public policies toward data sharing, user privacy, and killer data acquisitions affect competitive dynamics and efficiency. Finally, we show conditions under which a consumer coordination problem arises endogenously from data-enabled learning.

In recent years, the unique opportunities and challenges the firms face in the digital context and how the existing strategy concerns could be applied and expanded to accommodate such challenges have started to receive closer attention from scholars. More broadly, digitization calls for a more nuanced understanding of its implications on competitive and nonmarket strategies, but the conversation between the two sides has remained less integrated. This symposium attempts to bridge the dialogue between the market and nonmarket strategies, focusing on the digitally-mediated and enhanced interactions within firms (e.g., algorithmic decision-making, platform discoverability) and the political strategies of firms to navigate the heightened societal/regulatory pressures arising from such interactions. Algorithmic Predictions, Confidence, and Reasoning Quality Author: Xi Kang; Vanderbilt U. Author: Hyunjin Kim; INSEAD Optimal Discoverability on Platforms Author: Andrei Hagiu; Boston U. Author: Julian Wright; National U. of Singapore The Non-market Effects of Market Power: Evidence from Mergers and Lobbying Author: Bo Cowgill; Columbia Business School Author: Andrea Prat; Columbia Business School Author: Tommasso Valletti; Imperial College Business School Negative Lobbying by Digital Firms and Effects on Competitors Author: Angela Soomin Ryu; Columbia Business School

Platforms use price parity clauses to prevent sellers setting lower prices when selling through other channels. They claim these restraints are needed so platforms have incentives to invest in providing search services—without them, consumers would search on the platform but then switch to buy in a cheaper channel. In a model incorporating these effects, we find that wide price parity clauses lead to excessive platform investment while narrow (or no) price parity clauses lead to insufficient platform investment. Taking these investment effects into account, wide price parity clauses lower consumer surplus but have ambiguous effects on total welfare.

A growing number of digital platforms operate in a dual mode: running marketplaces for third-party products, while selling their own products on those marketplaces. We build a model to explore the implications of this controversial practice. We analyze the tradeoffs that arise from a regulatory ban on the dual mode, showing how such a ban can harm consumer surplus and welfare even when the platform would otherwise engage in product imitation and self-preferencing. In the empirically most relevant scenarios, policies that prevent platform imitation and self-preferencing generate better outcomes than an outright ban on the dual mode.

We explore the implications of steering by an informed profit-maximizing intermediary. The intermediary steers consumers by recommending firms, taking into account both the commissions firms offer and the prices they set. Such steering results in higher commissions and consumer prices, so that consumers only benefit from intermediation when their search cost is sufficiently high. Steering reverses the normal relationship between competition and price, with prices increasing in the number of competing firms. We use the framework to study various policies including commission caps (absolute or relative), commission disclosure, promoting information provision, and penalties for inappropriate advice. (JEL D11, D12, D82, D83)

We consider the reasons why a monopoly multi-sided platform may price differently from a social planner. The existing literature has focused only on the classical market power distortion and a distortion in the spirit of Spence. We show two additional distortions appear in the presence of cross-group network effects, which we call the displacement distortion and the scale distortion. We show conditions under which the displacement distortion exactly offsets the Spence distortion, and provide an example in which the total of these different distortions results in monopoly prices per user that are lower than the social planner's on both sides. Our results have implications for regulatory policy, which we briefly discuss. (C) 2021 Elsevier B.V. All rights reserved.

We update citation-based rankings of economics journals to study the relative ranking of new society journals. We employ two ranking methods: a standard iterative eigenfactor methodology adjusted for reference intensity and a novel top-5 citation alternative. We find that the American Economic Association journals ( AEJ-Applied , AEJ-Macro , AEJ-Micro and AEJ-Policy ) and the Econometric Society journals ( Quantitative Economics and Theoretical Economics ), are the top-ranked within their respective fields