🤖 AI Summary
This paper investigates information transmission and trust formation under “stealth learning” constraints—where senders cannot fully conceal or distort acquired information. Using game-theoretic modeling and Bayesian equilibrium analysis, it characterizes optimal sender strategies under bounded misreporting capability. The central finding is that senders influence receivers not through outright deception, but via strategic “selective ignorance”; the mere possibility of deception acts as a pivotal constraint shaping information acquisition and disclosure decisions. The paper establishes, for the first time, that enhanced verifiability is formally equivalent to strengthened commitment power, and rigorously characterizes both sender-optimal and receiver-optimal falsification environments. Key contributions include: (i) proving the endogeneity of strategic ignorance in no-deception equilibria; (ii) deriving necessary and sufficient sign-and-magnitude conditions for how increased verifiability affects sender utility; and (iii) fully characterizing the structural properties of optimal falsification environments for both parties.
📝 Abstract
I study a model of information acquisition and transmission in which the sender's ability to misreport her findings is limited. In equilibrium, the sender only influences the receiver by choosing to remain selectively ignorant, rather than by deceiving her about the discoveries. Although deception does not occur, I highlight how deception possibilities determine what information the sender chooses to acquire and transmit. I then turn to comparative statics, characterizing in which sense the sender benefits from her claims being more verifiable, showing this is akin to increasing her commitment power. Finally, I characterize sender- and receiver-optimal falsification environments.