How does short-form design amplify self-control problems in digital media? That is the question addressed in a new research paper by Renjie Bao of Princeton University, believed to be a Princeton job market candidate, highlighted by Marginal REVOLUTION (via Quan Le).

The paper, Just One More Clip, studies how the structure of short-form content creates persistent overconsumption. Its central mechanism: short units repeatedly renew temptation that lasts longer than each individual unit, which turns what would otherwise be a series of local temptations into sustained overconsumption. This builds on a long literature in behavioral economics, dating back to work on present bias and time-inconsistent preferences, which has long suggested that people plan to consume less of a tempting good than they end up consuming when the moment of choice arrives.

To test this empirically, the author uses microdata from a U.S. short-drama platform. Short dramas are serialized video content distributed in very brief episodes, typically around one minute each, often delivered through apps that charge users to unlock additional installments. The genre has grown into a sizable industry in recent years, with apps such as ReelShort and DramaBox competing alongside TikTok-style feeds for mobile viewing time. The analysis exploits a nonlinear top-up menu on the platform to infer users' intended viewing plans, allowing a comparison between what viewers planned to watch and what they actually watched.

The core finding is substantial: paying users watch 82.1% more than intended.

The paper's structural estimates imply an average temptation horizon of 11.2 minutes. That horizon is short relative to the full length of a drama, but long relative to individual one-minute episodes, which is precisely the condition under which short units can repeatedly renew temptation and produce overconsumption.

Counterfactual simulations in the paper show that larger decision units, default limits, and breaks improve long-run welfare. In other words, design choices, such as how content is bundled and how consumption is punctuated, can meaningfully mitigate the self-control costs that short-form formats impose on users. A calibration applied to short-video platforms more broadly highlights the wider welfare relevance of these findings. The question of how platforms should weigh user engagement against user welfare has drawn growing attention from researchers and policymakers, and findings like these, which quantify the gap between planned and actual consumption, provide empirical grounding for that debate over defaults and design mandates. What bears watching is whether platforms incorporate such mechanisms, such as bundling or default breaks, into their products.

The post was surfaced via Quan Le on X.