A Simplified Quest for Knowledge

Abstract

This paper develops a transparent, simplified version of Carnehl and Schneider (2025)’s model of knowledge creation. Our tractable framework, which yields closed-form solutions for key welfare trade-offs, preserves the essential economic mechanisms while eliminating mathematical complexity. We derive four main insights. First, contrary to the original model's emphasis on ``moonshots," our analysis explicitly demonstrates that expanding knowledge and then deepening it (the moonshot approach) is never socially optimal under direct welfare comparisons. The original model's case for moonshots relies on second-best arguments involving research costs and dynamic externalities, not on direct welfare considerations. Second, we identify a novel misalignment between private and social incentives in multidisciplinary research contexts. Even without research costs — where the original model predicts perfect alignment — researchers bridging large knowledge gaps between disciplines choose locations that create suboptimal knowledge structures. Third, we analyse how citation-based incentive systems affect knowledge creation trajectories. We show that systems that privilege unique contributions over shared ones align private behaviour with social welfare objectives, while those that reward shared contributions lead to excessive knowledge deepening. Fourth, our analysis provides precise characterisations of optimal knowledge creation paths under various initial conditions and offers clear guidance for science policy. By clarifying when interventions can address misalignments between researchers' incentives and social welfare, our simplified model offers practical insights for the design of research funding mechanisms.