A Statistical Analysis of Privacy Policy Design

Abstract

Privacy policies are essential to the notice-and-choice approach to online privacy in the United States. And yet these policies are unreadable, inconspicuous, and hard to understand. And, as I discuss in other scholarship, privacy notices are designed in ways that make their substance even harder to comprehend. This Essay takes the next step in a research agenda on privacy policy design. Based on a survey of Internet users, this Essay shows that when faced with a choice between a website with a privacy policy with, on the one hand, an aesthetically pleasing design and invasive data use practices, and, on the other hand, a traditionally-designed policy with strong data protections, majorities of users perceive that the website with a well-designed privacy policy will better protect their privacy. In addition, using statistical modeling, this Essay shows that greater knowledge of the legal implications of privacy policies is the only significant predictor of a user's capacity to discern strong privacy protections despite policy design. The Essay concludes with a discussion of the importance of public education and awareness in addition to improvements in policy readability.

Keywords: privacy, privacy policies, notice and choice, statistical analysis, empirical analysis, social science, law and society, law and social science, disclosure, sharing, design