Market power in the digital economy

Abstract

The rise of large technology firms is closely tied to the economics of information goods: high fixed and low marginal costs, strong network effects, and substantial switching costs. These features are not unique to digital markets, but converge there with exceptional strength. Successful firms attract more users, which makes their product more attractive and entrenches their position further, while scalability and global internet access leave scale economies barely constrained by geography. Competition therefore tends toward winner-take-most outcomes. These firms are no longer confined to the information sector, holding dominant positions in a wide range of sectors, where they act as gatekeepers to demand. Market power is setting prices above marginal cost or imposing unfavorable terms because buyers lack alternatives. Some degree of it can for instance finance innovation. But entrenched market power is linked to higher prices, weaker competition, and a shift of income from labor to capital. While its causes and consequences reach well beyond the economic domain, understanding the underlying economic mechanisms is essential for competition policy and regulation. Chapter 2 asks whether firms producing information technology differ systematically from non-IT firms in market power, measured through markups. Existing work documents rising markups across US listed firms since 1980 without identifying a sectoral driver, while theory predicts elevated market power specifically among technology-intensive firms. Standard industry classifications obscure this: Amazon is classified as retail and Uber as taxi services, though both are fundamentally software and data firms. The chapter develops a firm-level classification using natural language processing that captures them as IT firms. Applying it, markups of IT firms rise and diverge sharply from non-IT firms after 1996, coinciding with the diffusion of the internet. High market power characterizes technology firms more broadly, not merely a handful of dominant platforms. Chapter 3 studies how platforms build market power through business stealing: existing demand is redistributed toward participants, shrinking non-participants' share and deepening dependence on the platform. Exploiting the staggered rollout of Thuisbezorgd.nl across Dutch towns between 2011 and 2017, and combining administrative restaurant data with web-scraped participation data in a difference-in-differences design, the chapter finds that entry by out-of-town restaurants shrinks the local restaurant sector by 16–21 percent in hours worked. A rough estimate of between half and nearly all of participants' growth comes at non-participants' expense. Consistent with the theory, commissions that restaurant need to pay to the platform rose from 4 percent of total order costs in 2007 to 14 percent in 2022. Chapter 4 decomposes changes in the aggregate labor share between 2000 and 2017 in the US into within- and between-sector components, to illuminate the impact of sector-level changes on the aggregate labor share. Within-sector changes dominate the overall decline, while between-sector shifts have almost no net effect. However, decomposing that near-zero effect reveals substantial offsetting movements: sector-level changes in total factor productivity and capital intensity contributed positively, whereas the reallocation of labor between sectors worked in the opposite direction.