Monthly Archives: July 2014

Emergency Communication after Virginia Tech: Legitimising Laws and Policies

The Virginia Tech shooting provoked a number of ‘crisis’ or ‘emergency’ communication measures being implemented in colleges and universities, and also led to changes in federal law. In the last blog, I documented the emergency communication problems highlighted by the Virginia Tech shooting: the delay in emailing the emergency alert to staff and students; the vague nature of the first alert sent out, coupled with it failing to really inject any urgency into the message; the inconsistency between the institutional emergency management plan and emergency communication policies about the parties with the relevant authority to issue emergency alerts. In the next two blogs, I intend to take this further by looking at changes to federal and state laws, Virginia Tech University itself and how this created a market for the emergence of a mobile phone safety application.

Analysing the need to ‘take action’ after the Virginia Tech shooting brings to mind the argument of Power about secondary risk management, where institutional responses are guided by “cultural demands for control, accountability and responsibility attribution.” The lawsuit which was filed against Virginia Tech University following the delay in communication highlighted the need for accountability and responsibility. Notably, the delay in communication on April 16th 2007 was said to violate the ‘Jeanne Clery Disclosure of Campus Security Policy and Campus Crime Statistics Act’ — commonly known as the ‘Clery Act’ — (1991) applying to all colleges and universities that participate in federal financial aid programs. The purpose of the law was for transparency around crimes occurring on colleges and universities, by keeping a public crime log, publishing an annual crime report and disclosing security policies. More specially relating to the Virginia Tech incident, the ‘Clery Act’ (1991) requires that higher education institutions provide ‘timely warnings’ in emergency situations posing a threat to students and staff. As it transpired, the federal-level ‘Clery Act’ (1991) was amended in 2008 to include changes to emergency communication: notification is now required to be immediate, unless to do so would impede efforts to resolve the situation somehow; notification methods are to be publicised to staff and students, plus tested and annually reported; the authority of campus law enforcement to issue emergency messages is to be clearly defined in institutional policies. In terms of measurable effects, research conducted by Campus Safety Magazine found that, a year after Virginia Tech, 73% of survey respondents had bought or intended to purchase mass emergency notification systems. A survey by Rasmussen and Johnson found that 75% of respondents intended to incorporate mobile phone technology in their systems after the Virginia Tech shooting, compared to the 5% who had it in place prior to the incident.

At the state-level, the Virginia bill ‘First warning and emergency notification system’ (2008) mandated that emergency notification for public institutions of higher education be ‘comprehensive prompt, and reliable,’ with various communication mechanisms and the appointment of authorised, trained individuals to activate systems. In the case of Virginia Tech University institutional policies, the recommendations and subsequent changes made appear to be ‘corrective action,’ where the university sought to repair its image by reporting plans to improve its emergency communication system. In terms of improving mobile phone signals in buildings on campus, service providers sent over technicians to improve tower capacity. Sirens were also installed at two additional locations throughout the campus. Most notably, at the time of the shooting, VTU was in the process of updating the notification system, but the event — and perhaps the accompanying controversy and lawsuits about the delay in emergency communication — meant the process was expedited and so ‘VT Phone Alerts’ was fully installed and configured by July 2007. This system allows for users to either opt-in or out of subscribing to the system. Those opting in choose three contact methods ranked in order of preference: text messages, instant messages, emails (including non-university addresses), phone calls to office/residence, phone call to mobile phones, and phone calls to elsewhere (e.g. parents’ numbers). At the time of August 2007, almost eleven thousand students, faculty and staff had subscribed to the ‘VT Phone Alerts’ system. Further changes have been implemented since then. In 2008, digital signs linking with the ‘VT Phone Alerts’ system were added to classrooms; the year of 2009 saw the introduction of the ‘VT Desktops Alerts’ system sending a message to laptops/computers connected to the internet, both on and off campus; the system was expanded to six regions in Virginia in 2010. The most recent notification system consists of contacting mobile phones, telephones and instant messaging supplements, other systems of emails, desktop alerts sirens/loudspeakers, hotline, website, electronic message boards. Under these circumstances, policies have a ‘legitimising’ function outlining the course of action needed to fix the problem, where the implementation of these gives authority to the decision-making body.

[This blog was put together by analysing legislative documents and the institutional policies of Virginia Tech University, as well as other studies and academic scholars. It was first published in my doctoral thesis and will likely appear in future publications.]

Emergency Communication Problems Highlighted by the Virginia Tech Shooting

The 2008 Virginia Tech shooting resulted in a transformation of emergency communication procedures by highlighting existing flaws in institutional polices and federal law — the changes made will be the subject of future blog posts. For this blog, I wish to set the scene by explicating exactly what went wrong on 16th April, 2007, when a school shooting attack occurred at Virginia Polytechnic Institute and State University — this will be shorted to Virginia Tech University for ease of reading.

The inquiry ordered by the Virginia Governor, Tim Kaine, into the actions of Virginia Tech University officials circumscribed the mistakes made on the day of the shooting. One of the major problems highlighted by VTRP (2007/2009) was there were inconsistencies between the institutional emergency communication policy, which listed the Policy Group (VTPG) and Police Department (VTPD) as being authorised parties for releasing emergency alerts, and the emergency management plan, denoting that the campus police department, despite having the authority to develop or request an alert, had to wait for the university’s policy group to deliberate and then either approve or reject the request — the latter option was the one followed on the day of the shooting, as campus police did not have the computer code needed to issue an emergency message campus-wide.

Another problem was there were no templates of messages for different scenarios when such a measure would have allowed for a more expeditious process of sending alerts. Virginia Tech University’s emergency management plans prior to the shooting centred on natural disasters (weather problems and fires) and acts of terrorism in a post 9/11 context, with varying levels of severity ranging from zero to three. Surprisingly perhaps, given it occurred in a post-Columbine environment, there were no provisions in place for a school shooting scenario — this could perhaps be attributable to the fact that most school shootings until that point had occurred in middle and high schools rather than colleges and universities. Consequently, the content of the emergency notification sent to staff and students at 9:26am was not very specific or urgent in its warnings:

A shooting incident occurred at West Ambler Johnston [dorm] earlier this morning. Police are on the scene and are investigating. The university community is urged to be cautious and are asked to contact Virginia Tech Police if you observe anything suspicious or with information on the case. Contact Virginia Tech Police at 231–6411. Stay tuned to the [webpage] http://www.vt.edu. We will post as soon as we have more information.

The second message sent at 9:50am, ten minutes after the perpetrator had begun shooting again, this time in Norris Hall, was more specific and direct: “A gunman is loose on campus. Stay in buildings until further notice. Stay away from all windows.” As it transpired, the VTRP (2007/2009) investigators concluded that if the emergency alert had been more urgent or advised specific actions, this could have alerted more people to suspicious activities and hence enhanced their chances of survival.

Notably, it could be argued that the most severe mistake was the delay in sending the message. In the VTRP (2007/2009) documents, a timeline of events shows that, after the dorm shooting, Virginia Tech University engaged in actions such as locking down its Center for Professional and Continuing Education and cancelling bank deposits for that day. At 8:45am, a member of the Policy Group emailed a colleague in Richmond stating “Gunman on the loose. This is not releasable yet” and, four minutes later, sent another one saying “just try to make sure it doesn’t get out.” Schools in nearby Blacksburg went into lockdown at 8:52am. Despite this, the first alert to students and staff at Virginia Tech was sent at 9:26am, twenty-one minutes after the first period started; although a technical error delayed the message for a further fifteen minutes so it was actually sent at about 9:10am. The review panel members surmised that had the message been released before 8:30am, this could then have been possibly received by students and staff before they left for the first class of the day, given most of the people would already be in class or walking to class at 9:26am and thus unlikely to have the opportunity or even the need to check their email accounts.

As it transpired, the official review came to the conclusion that the high death toll of the Virginia Tech shooting was attributable to the delay in communication and that the Policy Group and, to a lesser extent, the campus Police Department, are responsible for this. The authors of the review do acknowledge, however, that Cho is highly likely to have carried out a mass shooting that day anyway given his intentions to do so; it might have been elsewhere if the university plan had been thwarted. This chain of events resulted in Virginia Tech University changing its institutional policies, updating tits technologies and a lawsuit filed by some of the relatives of those killed — these will all be discussed in future blogs.

[Material from this blog originally appeared in a chapter in the book Reframing Punishment: Silencing, Dehumanisation and the Way Forward. Future blogs will expand on this, by noting the policy changes to institutions, law and technologies after the Virginia Tech shooting.]

Threat Assessment Model: Online Threats

The last blog spoke about using behaviour in everyday life as part of threat assessment. Internet sites can also be auspicious locations for tracking the existence of potential threats. What can be taken from analyses of online comments is the importance of language and descriptiveness when assessing potential threats on there. Clearly, YouTube videos geared towards empathising and/or idolising school shooters is going to attract a certain kind of audience — these are, therefore, more likely to contain potential threats. Evident in the findings here is that a certain number of factors indicate the possibility of a potential threat.

 

-Looking at the key agents in the statements. Who are they? Upon whom is blame and responsibility being placed? A tendency to exonerate shooters from blame and place culpability with the victims, those in the school, and society as a whole show at the very least a willingness to have some degree of empathy with shooters. This, obviously, does not necessarily mean that the user expressing those comments is a potential threat, so this needs to be in combination with the other factors listed.

-Material that fits under the rubric of the ‘revenge and bullying thesis,’ where users adopt the stance that the victims deserved it, the perpetrator was achieving justice, and that bullying was the main driving force behind the attack. Here, the shooters are being ‘romanticised’ and their attack has created, to an individual susceptible to its influence, a ‘culture script’ of action as a way to resolve their problems. What makes this have the potential to be a threat is combining this frame with statements sharing personal experience of bullying, followed by expressions of a desire to also carry out a school shooting.

-The degree of attack planning, such as sharing knowledge of ways to gain weapons legally and illegally or naming specific targets, times and dates. When users are as detailed as this, it suggests something more than just an internet troll or bored kid looking for attention, as they would be more inclined to just put “I want to carry out a similar attack,” which is very vague, has an absence of a target and attack plan, and has not considered the means to obtain resources needed to execute a shooting.

-Statements paralleling past shooters or expressing narcissistic tendencies. The former would be in the form of “everyone else is to blame for making me want to do this,” which denotes persecution and a perceived injustice, and switches agency from the potential threat-maker to other (perhaps unspecified) agents. The narcissistic statements would fit with the traits outlined in earlier blogs, with comments like “I am so much better than everyone else and I laugh at their incompetence” if this was combined with other components outlined in this model. The over-reaction aspect is perhaps less of an issue on YouTube and other internet sites because the very nature of such debates means that discussions can commonly degenerate into rants and abusive comments — of course, this only serves to worsen the problem and any negative feelings the user already has.

-What should particularly be flagged are any discussions where users appear to be encouraging each other or possibly collaborating to plan an attack together. The more detailed and descriptive the discussion, the more likely it is that this could be shaping up to become something else. These should be closely monitored and interventionary action taken if required.

 

[This blog was put together by analysing comments on YouTube videos about school shooters. It builds upon the threat assessment material detailed in previous blog postings and is best used in collaboration with the offline behaviour threat assessment model proposed.]

A Threat Assessment Model: Offline Behaviour

The blogs published on the 4th, 8th and 11th June documented the personality characteristics of past school shooters: fragile male identities, a specific form of overt narcissism and fitting the state of ‘egoism’ conceptualised by the sociologist Emile Durkheim. Previous school shooters have shared the following characteristics: feeling a sense of injustice at the world and seemingly blamed everyone else for this; persuaded to carry out their attacks, either by specific persons, groups or predicated on the feeling that they had suffered throughout their lives; blame was bestowed upon everyone else bar the perpetrator for all their problems, tying in with the sense of injustice and persecution they feel; a lack of romantic success exacerbating matters, sometimes leading to stalking and harassment of females; excessive individualism, where the perpetrators felt they were ostracised by others and a lack of connection to society. It seems that the homicide-suicide of school shootings could encourage those with fragile, narcissistic identities in a state of excessive individualism to go through with the violent fantasies in their minds. Moreover, they know that once they have gone through with the murders, the time will come where law enforcement either shoots them dead or arrests them; therefore, they go into the rampages with the clear intention of killing themselves at the end.

With this in mind, a threat assessment model can be developed to be applied to both offline, everyday behaviours. Applying overt narcissism traits — not covert narcissistic characteristics, given they tend to revolve around a general sense of hopelessness and despair, not any active plans to sustain a high sense of self-esteem and demonstrate superiority — to warning signs could be indicative of a potential case for threat assessment: over-reaction (commonly aggressive or passive-aggressive) to the slightest criticism, high self-esteem that needs constant validation; the desire to be infamous and extreme fantasies of success and power, delusions of grandeur; a feeling of superiority combined with a sense of worthlessness; a sense of isolation from others in a particular environment and/or society in general. Taking all this into consideration, it is advised that threat assessments take into account the factors the narcissism and egoism factors outlined above when investigating a potential threat. It is of key importance hereto avoid ‘profiling,’ given this can lead to ‘false positives’ (people who fit the profile but have no intention of carrying out a school shooting) and ‘false negatives’ (where there is no evidence of the traits, but someone has the intention to perpetrate an attack). It is when the traits outlined above are combined with more disconcerting aspects, like fantasising about having power over others, expressing the desire to harm people in the school and intensive shooting practice, that red flags should be raised. The next blog will explore how this model can be coupled with online threats to create a hopefully more robust and thorough model of interrogating potential threats for their harm potential.

[Interested readers are directed to the blogs published on the 4th, 8th and 11th June for further information. A more detailed version of this model will appear in a chapter co-written with Dr. O’Grady to appear in the edited volume Gun Violence in American Society.]

Online ‘Leakage’: Interrogating Internet Activities as Part of Threat Assessment

The last two blog discussed school shooter admirers expressing their thoughts in the site YouTube. What is of key concern here is that users with similar mindsets sympathising with and admiring school shooters might ‘discover’ each other on social media and plan an attack together or perhaps encourage others to perpetrate a shooting — with this in mind, it certainly seems that online sites should be subject to more intense scrutiny and threat assessment.

The study conducted by Pollak et al. examining a series of school shooting incident for patterns found that in eighty-one percent of actual and planned incidents at least one person knew about the attacker’s plans beforehand. It was discovered that the most commonly cited reasons for bystanders not coming forward with threats were: not having a supportive school climate, lack of a positive adult influence, not taking the threat seriously, and having a close relationship with the threat-maker. In the case of YouTube comments, the ‘bystanders’ — other YouTube users and ‘guests’ who view the site but are not logged on — are in a much better position to report threats: they do not have a relationship with the threat-maker; even if some users do not take the threats seriously, others will; unlike a school environment, they do not need to worry about being scolded if it transpires that they ‘overreacted’ to the threat.
The counterargument could be made that users attracted to a particular kind of video (say, a school shooter fan video) would be more likely to encourage and desire a school shooting than report threats. This highlights the need for researchers in relevant fields (criminology, psychology, sociology) to prioritise further studies into this area and for online Intel to then feed its way through to threat assessment. A way to facilitate this process is for websites like YouTube to have links to the appropriate authorities for the countries its users are from; this would be more encouraging to bystanders, who may feel they do not know the most relevant organization to report threats to. Of course, the predicament here is that trying to find the ‘right’ authority that has jurisdiction over threats posted on the internet is problematic, as it draws resources away from other matters. The development of a threat assessment model could negate this to some extent, especially if partnerships could be fostered between the relevant bodies such as researchers, schools, local law enforcement.

One problem arising, however, is trying to conduct threat assessment on the basis of the threat alone does not reveal whether the threat-maker has the ability, motive and desire to actually carry out the threat which they speak. On sites like YouTube, no other information is generally known about users outside of the threat, bar that which they do reveal — it is questionable how much of this is true though — so this makes threat assessment using online Intel only a tougher process than piecing together someone’s history in a school environment. The context is which a threat is made is of key important here, as it reflects the threat-maker’s mental and emotional state at that specific moment in time; hence, a threat could be triggered by substances or a devastating event when the threat-maker actually has no intention of going through with it. Accordingly, this is even more of an issue with online threats, for some commentators, rather than actually planning an attack, will be internet trolls or even just bored children looking for attention or deliberately trying to start an argument.
A short-term solution to the threat requires the cooperation of the website in question to give some details about the user’s location so that schools in close proximity can be warned. Once the website has provided some information about the threat-maker, a long-term solution would need a partnership between researchers assessing online Intel and the local school board to see if there are any parallels between threats reported both by bystanders online and within schools. If users’ identities were narrowed down further, then that would require communication between researchers and individual schools to try and trace the individual — students at each school could be encouraged to report anything they see online to facilitate that process. This calls for the creation of a separate agency to monitor and investigate threats posted via the Internet, or for perhaps passing some legislation to deal with online threats (a number of states have included cyber-bullying in their bullying protocols, which is a good start.

[Threat assessment knowledge was used to put together this post. The next blog will document a threat assessment model using all the postings throughout June 2014.]