Endogenous Deterrence Costs

Posted by Social Science Research Network

Endogenous Deterrence Costs Alex Parkinson (Boies, Schiller, and Flexner LLP)

Abstract: There is an optimal rate at which actors that indisputably break the law will be detected, prosecuted, and punished. Legal regimes optimize deterrence by calibrating three inputs: the probability of detection, the rate of prosecution, and the magnitude of punishment. Optimal-law-enforcement literature casts these three inputs as static, exogenously controlled variables, each of which is capable of being carefully manipulated — even titrated — independent of the others. This Article problematizes the singular hegemony of this traditional account. Contra extant optimal-law enforcement literature, the three foregoing deterrence inputs can (and often do) exert endogenous pressures on one another. To advance this thesis, I present a stylized account of endogenous deterrence regimes. These regimes principally manifest when legal regimes cycle the magnitude-of-punishment input for a violation of a given law into the probability-of-detection input for that same law. Texture is added to this stylized account by drawing on several examples of endogenous deterrence regimes in both private and public law. This Article’s second-order contention is that these endogenous deterrence regimes can generate considerable costs. Specifically, endogenous deterrence regimes produce systematic deterrence failures, exact an independent social-welfare loss, and allow unelected, unaccountable actors to inflict agency costs by altering legal regimes absent any oversight, “making law” and extracting rents in the process. While effects these are not inevitable — endogenous deterrence regimes can also be deployed as neat institutional-design features that yield considerable gains — they are likely in the status quo. Endogenous deterrence regimes are systematically more likely to prove costly merely by virtue of the fact that so little attention has been drawn to their existence and effects.