The potential harms of the Tor anonymity network cluster disproportionately in free countries.

The Tor anonymity network allows users to protect their privacy and circumvent censorship restrictions but also shields those distributing child abuse content, selling or buying illicit drugs, or sharing malware online. Using data collected from Tor entry nodes, we provide an estimation of the proportion of Tor network users that likely employ the network in putatively good or bad ways. Overall, on an average country/day, ∼6.7% of Tor network users connect to Onion/Hidden Services that are disproportionately used for illicit purposes. We also show that the likely balance of beneficial and malicious use of Tor is unevenly spread globally and systematically varies based upon a country's political conditions. In particular, using Freedom House's coding and terminological classifications, the proportion of often illicit Onion/Hidden Services use is more prevalent (∼7.8%) in "free" countries than in either "partially free" (∼6.7%) or "not free" regimes (∼4.8%).

The Benefits and Costs of Cybersecurity Risk Reduction: A Dynamic Extension of the Gordon and Loeb Model.

This article develops a dynamic extension of the classic model of cybersecurity investment formulated by Gordon and Loeb. In this dynamic model, results are influenced by the rate at which cybersecurity assets depreciate and the rate of return on investment. Depreciation costs are lower in the dynamic model than is implicitly assumed in the classic model, while the rate-of-return threshold is higher. On balance, the user cost of cybersecurity assets is lower in the dynamic model than is implicitly assumed in the classic model. This difference increases the economically efficient size of the cybersecurity system in value terms, increasing the efficient level of risk reduction.

The Case against Commercial Antivirus Software: Risk Homeostasis and Information Problems in Cybersecurity.

New cybersecurity technologies, such as commercial antivirus software (AV), sometimes fail to deliver on their promised benefits. This article develops and tests a revised version of risk homeostasis theory, which suggests that new cybersecurity technologies can sometimes have ill effects on security outcomes in the short run and little-to-no effect over the long run. It tests the preliminary plausibility of four predictions from the revised risk homeostasis theory using new survey data from 1,072 respondents. The estimations suggest the plausible operation of a number of risk homeostasis dynamics: (1) commercial AV users are significantly more likely to self-report a cybersecurity event in the past year than nonusers, even after correcting for potential reverse causality and informational mechanisms; (2) nonusers become somewhat less likely to self-report a cybersecurity event as the perceived riskiness of various e-mail-based behaviors increases, while commercial AV users do not; (3) the negative short-run effect of commercial AV use on cybersecurity outcomes fade over time at a predicted rate of about 7.03 percentage points per year of use; and (4) after five years of use, commercial AV users are statistically indistinguishable from nonusers in terms of their probability of self-reporting a cybersecurity event as perceptions of risky e-mail-based behaviors increase.

Drug transactions and the dark web: Public perceptions of the locational setting of offenders and support for drug policy outcomes.

BACKGROUND: Both legal and extra-legal factors influence judicial and non-judicial opinions about persons who use drugs. Yet, how the locational setting of drug transactions influences public perceptions of drug control policies remains understudied. In particular, the public's view of drug exchanges on the dark web could directly and indirectly influence drug policy, legal decision making, and spending decisions. The study's aim is to identify whether the location of a drug exchange, specifically the dark web, influences public preferences for drug policy and police resourcing. METHODS: A sample (n = 1359) from the United States of America was recruited and participated in a discrete choice experiment. The participants compared and repeatedly chose across five iterations between two drug offender profiles with nine set features, such as the location of drug transactions, all with randomized levels. The resulting sample included a total of 13,590 contest pairs. RESULTS: Averaging over the non-locational attributes, respondents indicated that, compared to the dark web, several locational settings for drug exchange (such as the street corner, social media, and an unknown location) needed fewer police resources and offenders were less deserving of longer punishments. No statistically significant difference was found for opinions about harm to communities, and offenders involved in drug exchanges on university campuses were considered more deserving of a substance abuse treatment program than offenders on the dark web. CONCLUSION: There appears to be a preference for more punitive criminal justice policies for drug transactions occurring on the dark web relative to some other common settings. Such preferences may indicate a novelty effect driven by negative sentiment surrounding the dark web or a perceived deficit in the police's ability to deal with drug crimes on the dark web. These findings suggest that the public may prefer supply-side policing efforts over demand-side policies, which emphasizes harm reduction.

The Dark Web and cannabis use in the United States: Evidence from a big data research design.

BACKGROUND: Cannabis is one of the most commonly sold drugs on cryptomarkets. Because of the anonymity-granting functions of Tor, no study has traced the within-country effect of the Dark Web on cannabis consumption patterns. This article uses a big data research design to examine the association between revealed interest in the Dark Web and self-reported cannabis use within US states from 2011 when Silk Road launched to 2015 when Operation Onymous shuttered nine markets. METHODS: This study uses mixed effects ordinary least squared regressions to analyze U.S. state/year panel data, using robust standard errors to correct for heteroscedasticity. Marginal effect plots illustrate substantive effects. The dataset consists of state-level variables drawn from the Uniform Crime Report (UCR), the American Community Survey (ACS), the National Survey on Drug Use and Health, the Correlates of State Policy Project, and the Bureau of Justice Statistics Justice Expenditure and Employment Extracts. Data for the Dark Web interest measure are drawn from Google Trends. The proxy for Dark Web interest is an index of eight Dark Web related search queries. RESULTS: The regression analysis indicates that Dark Web interest in US states positively correlates with cannabis consumption rates overall and among older adults (26+), but not youth (12-17) or younger adults (18-25). Additionally, Dark Web interest is positively associated with more frequent cannabis usage rates (i.e. use in the past month, excluding first time use) both overall and among older adults, but not among youth or younger adults. Dark Web interest does not correlate with casual use (i.e. use in the last year, excluding use in the past month) for any age bracket. Interacting Dark Web interest with state-level legalization regimes indicates that the association between Dark Web interest and cannabis consumption in the past year is no different in medically legalized states and amplified in states with recreational legalization. Lastly, the Dark Web interest term does not correlate with first time cannabis either overall or for any age category. CONCLUSIONS: Interest in the Dark Web is associated with increased cannabis use in U.S. states from 2011-2015, but the effect is concentrated in states with more frequent cannabis users, older users, and in states with recreational legalization of cannabis.