Some Notes on the University of New Hampshire Sextortion Report

As Ben noted last week, Janis Wolak and David Finkelhor of the University of New Hampshire’s Crimes Against Children Research Center recently released a new report on sextortion, funded by Thorn. I have now read it carefully. The report and our two papers are useful complements to each other in sketching out the universe of sextortion crimes. Below are a few notes on the UNH report and what can be gained from comparing our respective research.

The UNH report uses a very different methodology from our study. Whereas we gathered data on sextortion by focusing on prosecuted criminal cases, searching LexisNexis and WestLaw for information, Wolak and Finkelhor conducted a survey of sextortion victims who came forward in response to advertisements on sites like Facebook and Twitter.

For this reason, both data sets are incomplete--though in different ways. In writing our report, we were reliant on court records and news reports, so any instance of sextortion that stopped short of entering the justice system necessarily wasn’t included in our data. And such instances seem to make up the majority of the UNH report, which records that only 16 percent of respondents reported the incident to the police. What’s more, only 21 percent even reported the situation to the website or app over which the abuse was taking place. (Memo to social media companies: making such reporting easier is something to work on.)

On the flipside, the data on sextortion in the UNH report was wholly dependent on those who both happened to know about and who chose to complete an advertised survey. Wolak and Finkelhor make this clear up-front, stating that “the respondents do not constitute a representative sample of persons who have been the targets of sextortion.”

For the same reason, the UNH report is also entirely dependent on the information the survey respondents provided, with no possibility of fact-checking. It’s entirely plausible that some of the stories told by survey respondents might not rise to the level of sextortion or merit criminal charges once entered into the justice system. So whereas our dataset was underinclusive, in the sense of including only cases in which prosecutors believed they could prove a criminal violation beyond a reasonable doubt, the UNH dataset may be overinclusive in the sense of including unverified allegations by anyone who claims to be a victim of sextortion. This is not to minimize the pain felt by the respondents to the UNH survey or to deny their experiences. Rather, it’s just to note that a victim’s feeling of suffering, and law enforcement’s ability to prosecute a case―even with the use of a sextortion-specific statute―will not always overlap.

The self-conscious limitations of both studies are a reminder of both the paucity of data available on sextortion and the difficulty of gathering sextortion data in the first instance.

A second important methodological difference involves the definitions of sextortion―a term without an agreed-upon definition―in the respective reports. Our reports defined sextortion as:

old-fashioned extortion or blackmail, carried out over a computer network, involving some threat—generally but not always a threat to release sexually-explicit images of the victim—if the victim does not engage in some form of further sexual activity.

In contrast, Wolak and Finkelhor’s report reads:

Sextortion is defined as threats to expose a sexual image in order to make a person do something or for other reasons, such as revenge or humiliation.

In other words, whereas we confined our definition of sextortion to cases in which the sextortionist requested further sexually explicit images or video from the victim, Wolak and Finkelhor include cases in which something else besides sexual material was extorted.

This difference is important because a small proportion (9 percent) of the cases included in the UNH report involved money as the item being extorted. A comparable percentage of victims (10 percent) reported being forced to hurt themselves through this extortion. While these stories are certainly appalling, they would not fall under our definition of sextortion.

On the other hand, because respondents to the UNH survey could select more than one option as to what was extorted from them, it’s conceivable that some of these sextortionists might have forced their victims to provide both sexual images and, for example, money. Such cases did not appear in our dataset, but that by no means indicates that such things don’t happen.

More strikingly, in a full 42 percent of Wolak and Finkelhor’s cases, the sextortionist sought to force the victim to either remain in a sexual relationship with him or return to such a relationship. Our dataset included some sextortion cases that could also be categorized as intimate abuse of this sort. In one case, for example, the perpetrator sextorted his former partner in an attempt to get her to meet and have sex with him; in another, a man threatened to make public sexual images of his ex-girlfriend unless she sent him additional images. Yet these instances made up only about 5 of the 78 sextortion cases analyzed in our report, versus almost half of Wolak and Finkelhor’s cases.

This is an enormous discrepancy, one that actually goes to the core of what the average sextortion case looks like. In the UNH study, intimate partner violence makes up nearly half of sextortion cases; in ours, such violence contributes a relatively small amount to the universe of cases. This is underlined by the UNH data on online versus offline cases: 59 percent of the respondents described a face-to-face relationship, of which 64 percent indicated that the extortionist was a current or former intimate partner.

I suspect that part of the explanation for this discrepancy is that while these cases might well fall under our definition of sextortion, they tend to get prosecuted as instances of domestic violence or stalking, and so are not easily swept up in our searches for sextortion cases. Another possible explanation is that large numbers of these cases are not being reported to the police at all: not only does intimate abuse often go unreported, but the UNH report suggests that victims were somewhat less likely to report extortion cases when the perpetrator was blackmailing them to remain or return to a relationship. A third possible explanation is that these cases are more likely to be handled by state and local law enforcement, and thus less likely to show up in our dataset.

Keeping in mind that Wolak and Finkelhor’s data involve a self-selected sample and thus a potentially unrepresentative group, their report includes some interesting tidbits concerning the demographics of both perpetrator and victim populations.

A significant number of their respondents were over 18 when their harassment began. As Ben noted, this is consistent with our own data and underlines the importance of addressing the sentencing discrepancy between sextortionists with underage versus adult victims.

Wolak and Finkelhor’s data regarding the gender of victims and perpetrators is noteworthy as well. As we did, they found a notable number of male victims, though the majority of victims were female. But their survey respondents also report that in 9 percent of cases in which the extortion occurred in face-to-face relationships, the perpetrator was female. This is a notable departure from our dataset, in which literally all perpetrators are male.

The UNH report also highlights the popularity of social media manipulation, or “catfishing,” among sextortionists. Of respondents who first met their harassers over the internet, about half reported being misled as to their identity. This is consistent with our findings as to the prevalence of catfishing as a method to obtain an initial explicit photo or video before beginning sextortion.

One major benefit of Wolak and Finkelhor’s approach and its focus on the experience of victims is that the report makes abundantly clear the devastating effects that sextortion has on those who survive it. As Ben noted, an astonishing 12 percent of those surveyed in the UNH study reported having to move to a new neighborhood or town as a result of sextortion. Twenty-four percent saw a medical or mental health professional. If any doubts remained regarding the severity of this problem, these statistics—along with the victim testimony that appears throughout the report—should quell it.

Finally, Wolak and Finkelhor’s data reinforces our point about the scale of this particular problem. They report responses from 1,631 victims of sextortion. In comparison, our report counted 1,397 victims identified by authorities as specifically targeted by sextortionists, with an estimate of the total number of victims in all our 78 cases as ranging anywhere from 3,000 to roughly 6,500.

Again, given the disparate nature of the data gathered in both reports, it would be premature to attempt to draw any hard conclusions from these statistics about sextortion’s prevalence—except one, that is: Sextortion happens a lot. If more than a thousand victims responded to Wolak and Finkelhor’s survey, more than a thousand were identified in the cases we reviewed, and an additional two to five thousand victims may also have been affected by the sextortionists prosecuted in our cases, we should all be asking ourselves: how many more sextortionists and how many more victims are out there?