Differentiated Regulation Across People and Situations: A Behavioral Ethics Perspective to Personalized Law

Abstract

Wrongdoing is ubiquitous. Media outlets constantly report an endless stream of deleterious human behavior, from sexual harassment and fraud in financial markets to corporate and political corruption. Recent developments in behavioral ethics research suggest that these ills will forever accompany human interaction due to the phenomenon of “bounded ethicality,” or people’s limited ability to conduct an objective and candid moral examination of their own actions. When evaluating the ethical implications of their behavior, individuals have been shown to be biased and to systematically underestimate or ignore the magnitude and effect of their own misconduct. Such findings have troubling implications from a law enforcement perspective. That is, if wrongdoers are able to convince themselves they are doing nothing wrong, how can regulators and policy makers ever successfully reduce or prevent misconduct? Essentially, recognizing the power of bounded ethicality reinforces the idea that destructive human behavior may be unavoidable, and that it may never be possible to reduce the systematic wrongdoing currently observed throughout society. In response to the challenges that bounded ethicality poses, this Article explores how using big data analytics contributes to curbing both ethical bias and the results of bounded ethicality. The paper is breaking new ground in being the first to explore the intersections between the growing literature on behavioral ethics, highlighting the concept of bounded ethicality, and the scholarship and research on data-driven law enforcement. We suggest that to combat bounded ethicality, regulators should use ethical nudges, regulatory interventions designed to improve ethical deliberation by potential wrongdoers. We show that the use of big data analytics is crucial for the successful deployment of such regulatory interventions, for several reasons. First, ethical nudges must be deployed in real time, when potential perpetrators are making their decisions regarding possibly unethical actions. Big data analysis can facilitate this type of timely regulatory response in its shift from ex post inquiry to ex ante prediction. By collecting and analyzing data on the antecedents of wrongdoing, big data analytics can enable regulators to respond at opportune moments and situations, rather than engage perpetrators ex post facto. Second, ethical nudges must be targeted rather than general. If individuals are nudged constantly, ethical nudges will lose their effectiveness due to the phenomenon of ethical numbing. By identifying the situations that breed unethicality and limiting nudges to those situations, big data analysis can ensure that people are not overexposed to ethical nudges, thereby maintaining their efficacy. Third, ethical nudges must also be tailored to the characteristics of the specific bias that is causing unethical behavior in each specific case. Using big data analytic, together with behavioral ethics insights, regulators can collect information that will indicate the nature of the ethical bias operating in specific instances of wrongdoing, and thus be able to deploy the appropriate regulatory response. In addition to presenting several other advantages of using big data analytics as part of the efforts to reduce bounded ethicality, this paper suggests a full menu of regulatory tools designed to improve moral deliberation and discusses the importance of big data analytics as a basis for their use. Our analysis calls for a reorientation of existing practices of data-driven law enforcement, to make it more suited for the regulation of bounded ethicality. We show that this reorientation is also necessary for reasons related to the legitimacy of data-driven law enforcement. Data-driven law enforcement currently adopts a personalized focus, attempting to identify individuals who are more likely to commit crimes. This approach is highly problematic, raising privacy concerns and perpetuating discriminatory practices. Conversely, the approach we advocate calls for a focus on situations, not individuals, as behavioral ethics studies show that bounded ethicality is primarily situation driven. Therefore, big data analytics should be used to identify situations that breed unethicality, thus shifting the focus to individuals more likely to act badly. This use of big data analytics is less harmful to individuals’ privacy, as it does not focus on personalizing legal treatments.