Automated Penalty Institutions on Staked Blockchain Networks
Abstract
Blockchain networks provide a novel form of economic organizational coordination through the protocol’s ability to facilitate joint production absent the traditional locus of centralized authority that characterizes the private firm. But the joint productive scope of this coordination is limited if it relies exclusively on rewards to incentivize good faith performance. This has led to the emergence of automated penalties, also called “slashing”, to better approximate the balance of incentives that firms have long relied upon to motivate employees to exert more effort. Yet just as firms’ specific contractual arrangements vary considerably depending on the joint production that they coordinate, staked blockchain networks’ penalties vary considerably in practice. We survey the largest staked networks by market capitalization to develop a comparative institutional analysis of their penalties, identifying the actions that typically trigger penalties, common classes of penalties, and means by which penalties are assessed. We detail reasons for this variation including contextual tailoring, random variation, mimicry, and default rules’ existence trumping specific rule choices. Finally, we consider the broader applicability of automated penalties in coordinative network contexts, including those on which numerous AI agents will be making risky decisions on behalf of a human principal.