Tor and the City: MSA-Level Correlates of Interest in Anonymous Web Browsing

Abstract

In the twenty-first century, Americans have put more of their lives online while the US government has expanded its mass surveillance apparatus. Interest in anonymity-granting technologies like The Onion Router (Tor) has grown substantially as citizens seek to protect their privacy. However, this same technology can be used to engage in illegal activity on the dark web. This study examines how interest in the dark web, public attention to the 2013 Snowden revelations, and metro-area political ideology are associated with public interest in Tor. We link data from multiple sources including Google Trends, the American Community Survey, the Bureau of Labor Statistics, and the Cooperative Congressional Election Study for the forty-nine largest US Metropolitan Statistical Areas (MSAs) from 2006–2015 (n=490). Broadly, we find that metro areas with liberal citizen ideology and greater interest in the dark web were more likely to search for Tor. When controlling for the level of interest in the dark web, the Snowden revelations of 2013 had no significant impact on interest in Tor. These findings suggest that the lure of the dark web and left-leaning ideological contexts offer stronger explanations for interest in anonymity-granting technology than the public attention brought to mass surveillance by the Snowden revelations.

Seeking anonymity on the Internet: The knowledge accumulation process and global usage of the Tor network

“…First, the study fills in the existing research gap between contextual motivators and eventual use of new media technology by examining the predictors of Tor usage levels from the perspective of the knowledge accumulation process. Furthermore, existing studies have only examined searching behavior as a predictor of the search for “Tor”-related topics or offline activities within the United States (Jardine and Lindner, 2020; Lindner et al, 2020; Lindner and Xiao, 2020). This study extends the outcome variable from the search for “Tor” to real usage in a cross-national setting.…”

“…For example, the search terms under the search topic “Censorship” include censorship (English), censura (Spanish), цензуры (Russian), روسناس (Persian), 審査 (Japanese), and so on. Although the exact search terms are not disclosed, search topics in Google Trends are generally considered to provide more comprehensive information compared with single search terms (Lindner et al, 2020). To systematically examine the potential related topics to Tor, we compiled terms from various previous studies for three types of knowledge accumulation (see Table 1), that is, the accumulation of “Tor”-related knowledge, “Digital Rights”-related knowledge, and “Drug Markets”-related knowledge, respectively.…”

“…“Tor”-related knowledge accumulation involved search topics on Tor, as well as on other highly related technologies, such as Tails and PGP, which provides higher-level security when using Tor (Kowalski et al, 2019). “Digital Rights”-related knowledge accumulation included search topics on “Surveillance” (Lindner et al, 2020; Lindner and Xiao, 2020) and “Censorship” to refer to the user interests of mass surveillance by governments and institutions, which are important factors that determine attitudes toward Tor (Jardine, 2018a). This cluster also included “Privacy,” “Private browsing,” and “Data breach” to cover users’ concerns about personal data leakage in the online environment, which are also significant drivers of using Tor (Jardine, 2018a).…”

“…A worldwide survey has demonstrated that concerns about privacy, censorship, and data breaches significantly predict attitudes toward the Tor network (Jardine, 2018a). Subnational studies further found that people from metropolitan areas with more liberal citizen ideology are more likely to become interested in Tor and search for related knowledge (Lindner et al, 2020). In particular, existing evidence in the United States at a state-year level has suggested that searches for topics related to Edward Snowden’s revelations are more potent for predicting the search popularity of Tor in comparison to searches for the Dark Web (Lindner and Xiao, 2020).…”