22 December 2013

Identity Scanners

'ID scanners in the night-time economy: Social sorting or social order?' (Australian Institute of Criminology Trends & issues in crime and criminal justice no. 466) by Darren Palmer, Ian Warren and Peter Miller [PDF] examines
the rationales for adopting compulsory patron ID scanning as a key method of reducing violence in and around licensed venues in the Victorian regional city of Geelong. Using a mixed methods approach, this paper challenges the popular perception that ID scanning has helped to reduce violence Geelong’s night-time economy. Further, the research identifies several limits in the administration of this technology that potentially undermine patron safety in the night-time economy. 
The authors comment that
Extensive media coverage and public outrage towards extreme cases of violence that lead to serious injury or death, frequently provide the impetus for introducing new and untested methods of enhancing venue security. One example involves the mandatory installation of high-resolution CCTV systems in all licensed venues throughout Victoria to increase patron surveillance and allow for the rapid identification of both perpetrators and victims of violence (Department of Justice 2009; VLRC 2010). However, evidence of the impact of CCTV in preventing, reducing or deterring crime remains equivocal (Wilson & Sutton 2003).
Digital identity authentication systems have similar appeal (Lyon 2009). ID scanning invokes ‘commonplace’ technology (Goold, Loader & Thumala 2010), such as a desktop or laptop computer, a digital camera, a portable scanning device and a hard drive, to automate conventional manual identity screening processes before a patron is granted entry into a venue. With appropriate software, these systems enable door staff to accurately identify and prevent entry to those with a previous record of violence at the venue.
At the point of entry into a licensed venue, a person must produce a driver’s licence or passport, or their fingerprint. An image of the document or biometric identifier is then scanned into a portable computer located at the venue or conveyed to a server operated by the system manufacturer. Most systems also require a digital photograph to be taken when a patron’s identity is first recorded and at each subsequent attempt to enter the venue. The technology then automatically matches the photograph with the person’s identification document or biometric fingerprint. System administrators can then manually adjust an individual’s digital record if they have engaged in any violent behaviour or are evicted from the premises. If a banned person attempts to re-enter the venue, automated matching of the compulsory photograph with the ‘flagged’ identity record enables door staff to readily identify those who should be denied entry.
Variants of this technology have been adopted to manage street offending in crime ‘hot spots’ (Beckett & Herbert 2008; Gibson 2003) and to facilitate penalties for the sale of alcohol to minors in the United States (Cross 2005). As with CCTV, the presence of an ID scanner might deter some patrons from attempting to enter a venue or engaging in violent and disorderly behaviour. Increasingly, data networking can also enable rapid digital information sharing between venue operators, security providers or police, which can assist in the enforcement of patron bans or prohibitions on underage entry into late-night venues.
There are currently few ID scanning systems available on the Australian market. This means it is easier for system developers to ‘share a banned list of troublemakers—whether that listing is local, statewide or national’ (O’Brien & Duff 2011: 4) among all venues employing the same system. Therefore, ID scanners have enormous potential to address ongoing concerns about security and law enforcement arrangements in the night-time economy, given the seemingly pervasive risks of violence in Australian drinking culture (Tomsen 2005).
The research has a strong criminological emphasis, with an emphasis on the need for more research and a disappointingly weak engagement with privacy issues such as those highlighted here and here. The authors comment that -
All but one venue licensee favoured the mandatory use of ID scanners and further data sharing across all ‘high-risk’ venues in the Geelong CBD. Suppliers of this technology, police, security personnel and local council officers endorsed these favourable views subject to minor qualifications.
Benefits of ID scanners
Respondents considered ID scanners to be the centrepiece of several measures under the revised 2007 Accord that sought to proactively ‘do something’ to improve the management of the Geelong night-time economy and reduce both the frequency and brutality of violent confrontations. One licensee indicated most Accord participants initially viewed ID scanners as a contentious policy option. However, their effectiveness at ‘high-risk’ venues could be seen through discernible shifts in patron behaviour after their implementation.
I was apprehensive at first but since I implemented them over two years ago it’s probably been one of the biggest tools that I believe has changed the behaviour of the patrons in our venues (Licensee).
ID scanners were also considered to reduce antisocial behaviour in licensed venues by accurately identifying people with a recorded history of disorderly conduct and ensuring patron bans could be readily enforced. This served two important deterrence functions. First, licensees believed potential troublemakers avoided attempting to enter venues where scanners were deployed. Second, the increased probability of accurate and rapid identification was considered to attract more orderly patrons. Both venue licensees and suppliers considered ID scanners removed ‘anonymity’ and reduced the likelihood that people with a propensity to engage in violence would attend the CBD nightclub precinct.
By removing anonymity, those who are prone to bad behaviour, and not necessarily just because of alcohol...but the fact that they know that they’re not anonymous, it’s pretty much a surety that they are going to be caught, be able to be identified and then caught, so they don’t do it. They’ll go elsewhere (Licensee).
Cost efficiency was significant, as ‘safety is business’ in the night-time economy. Both licensees and system suppliers indicated ID scanners are a low, one-off expense that is easy to install, maintain and upgrade. Most third-party system administrators provide low-cost technical support, software upgrades and data storage facilities. Thus, the technology is considered relatively easy to administer on-site and involves limited financial outlay.
Appearing to do something proactive about violence in the night-time economy, deterrence and cost efficiency were consistently viewed by venue licensees and other Accord participants as countering any perceived limitations of this technology. Commercial imperatives reinforced these themes among suppliers of ID scanning systems. Although one licensee expressed concern that police had placed undue pressure on ‘high-risk’ venue proprietors to implement the technology as a concession to the local liquor industry’s resistance to a proposed 2.00 am lockout at all venues, ID scanners were valued as a visible, deterrent-based and cost-effective measure designed to reduce violence. Only one licensee acknowledged that scanners might not deter drunken patrons engaged in ‘spur of the moment’ confrontations, who are unlikely to consider or care about any potential ramifications from identity-based or CCTV surveillance.
When questioned further, it became clear that venue licensees, door staff and security personnel were also aware of several anomalies in the use of ID scanners that could impact on, but not override, their potential to reduce violence. During peak times when long queues can lead to patron antagonism, a policy of selective (non)scanning was implemented at most larger venues. Patrons considered to be non-threatening, such as young women, were generally ushered past the scanning unit without undergoing an identity check. Young men who conformed to an accepted risk profile were commonly entered into the system as a matter of course.
Door staff questioned the uneven application of selective non-scanning. In line with recent observations in Edmonton, Canada (Haggerty & Tokar 2012), one door worker indicated this ‘no hands on’ policy undermined the potential for ID scanners to reduce violence. Friends of door and security staff, preferred customers and venue members routinely bypassed the system with no electronic record of their presence (Haggerty & Ericson 2000). One respondent indicated it was difficult to challenge this practice.
[T]he bouncers do let in people that they know, like bikies, really rough people, and those kinds of people can make everyone else feel intimidated…But there [were] always fights. It seemed to always kind of be the same kind of people...Like they kind of just went there for fighting…I don’t think it [a scanner] really makes a difference…I think they are going to do it [fight] anyway because most of the time they are wasted so they are not going to think about ‘the scanners are there, I’m not going to do this’ (Door staff).
While a perceived major benefit of ID scanning is its potential to remove patron anonymity, selective non-scanning undermines this in two ways. As the above quote indicates, it can enable security staff to allow patrons willing to engage in violence to enter licensed venues (Haggerty & Tokar 2012). However, by informally profiling young men deemed to be potential ‘troublemakers’, a considerable proportion of potential victims of violence might remain difficult to identify if their personal details are not entered into these systems (Harcourt 2007). This second issue was not recognised by any key stakeholders interviewed for this study.
All venues are required to display ‘clear signage at the front of the premise explaining that the patrons details were kept for 28 days and then destroyed’ (Licensee). However, door staff indicated there were few protocols for describing the ID scanning policy to concerned patrons. While this was rarely necessary, door staff commonly advised ‘the law’ mandated ID scanners. Concerned patrons could try entering another venue or were advised to contact system administrators during business hours. Concerns were seldom raised about whether this advice undermined a patron’s voluntary consent to having their personal information shared and stored at the venue.
Data retention policies are also adapted from principles developed for CCTV, requiring personal information to be deleted after 30 days. Suppliers indicated various operating procedures and technical protocols were developed in line with both state and federal privacy laws. One biometric system incorporated a complex data encryption system that could only be accessed by third-party administrators and was developed through the use of privacy consultants to ensure compliance with the national privacy principles. However, this system has yet to be incorporated in Geelong.
One ID system supplier questioned the lack of clear regulatory standards regarding the collection and dissemination of personal data. As demand increases for interoperable technology and greater information sharing about banned patrons among venue managers, police and security agencies, there is a need for the development of clear regulatory and data management protocols. Nevertheless, the following quote also highlights that increased regulation may prioritise information privacy over patron safety.
The fact that there are no protocols…leaves the use of these things vulnerable. They have been highly effective and the misuse of one could bring down a lot of good work and that worries me quite a bit. So in terms of is regulation needed for it? Yes. But the problem is when you get regulators involved…they are largely coming from an angle that is myopic and not a balance of what the real objective is; and that is making it safe. And making it safe means there has to be some surrendering of privacy (System Manufacturer).
Licensees favoured the open circulation of information between all Geelong venues and with police to prevent ‘bar hopping’ by flagged or banned individuals. As one venue licensee described:
If you can get it [ID scanning] implemented on a grand scale in the CBD at least it has some sort of impact because if they were to be banned from [one venue]…[unruly patrons] used to be able to walk into any other venue they can. But now it impacts on all the CBD venues so if they’re banned they’re banned everywhere…(Licensee).
However, this study revealed an important paradox associated with information privacy and crime prevention. Victoria Police consistently used information privacy as the standard justification for only providing generalised LGA data rather than the specific locations of reported assaults or basic demographic characteristics of victims and suspects. Local government representatives participating in the development of the November 2007 Accord considered this was the major impediment to the development of meaningful evidence-based policies to combat alcohol-related violence. Prior to calling for more research the authors comment that Despite claims by venue licensees and system suppliers, this study provides limited empirical support for claims alcohol-related assaults in and around Geelong’s late-night venues have declined since the initial pilot of May 2007 and the subsequent mandated use of ID scanners under the revised Liquor Accord. Between May 2007 and May 2008, there was no discernible reduction in either reported assaults (see Figure 2) or emergency department admissions that identify alcohol-consumption as a key variable. Increases in street assaults from July 2008 suggest ID scanners may have produced a displacement effect, but firm conclusions on this point are not possible. This is an example of one of Brown’s (2013) key principles of regulation of crime prevention interventions—they must be effective at preventing crime.
A different picture emerges when these figures are considered in light of key stakeholder interview data. Venue licensees and system managers equate improved business with the erosion of anonymity and deterrence. However, selective non-scanning has significant potential to undermine these key benefits. Door staff indicated this form of profiling had not reduced violence at some Geelong venues. More problematically, the inability to rapidly identify a significant proportion of potential crime victims is a recognised by-product of selective profiling (Harcourt 2007). Selectively targeting potentially ‘troublesome’ young men allows certain individuals or groups to remain anonymous and potentially less readily identified if a victim or offender. This generates a rather skewed ‘surveillant assemblage’ (Haggerty & Ericson 2000: 619) that is difficult for those who are proactively targeted to contest.
However, ID scanners have considerable value in enforcing bans from licensed venues. By nature, banning policies can only be enforced through identity authentication (Beckett & Herbert 2010; Lyon 2009). System suppliers, venue licensees and security staff unanimously agreed that the use of ID scanners is an efficient method for excluding banned patrons. Most Australian states have legislatively mandated short-term public order bans and long-term prohibitions on entering licensed venues or nightclub precincts for serious alcohol-related offences that operate alongside a licensee’s proprietary right to deny entry or evict any patron (see Palmer & Warren 2013). The information about patrons that have been banned by venues or have been subject to legislated banning orders is shared between police and venues via the local Liquor Accords, thus avoiding breaches of privacy regulations.
However, the lack of clear policy guidance on implementing ID scanning technology enables door staff and security personnel to subvert the intent of removing patron anonymity, which can occur at the expense of ‘the actual task of providing [human] protection’ (Zedner 2006: 277). In Harcourt’s (2007: 23–25) terms, this is a form of ‘elasticity’. Selective profiling that aims to enforce venue exclusions legitimises gaps in the administration of ID scanning that either allow other forms of crime to remain immune from the surveillant assemblage, or that potentially compromises the identification of victims of violence with no recorded electronic profile. This helps to explain why there has been no significant decline in reported assaults or emergency department admissions between November 2007 and mid-2009.
Crime prevention technologies present numerous challenges for privacy regulators in Australia (ALRC 2008) and internationally (Brogan 2002/2003; Goold & Neyland 2009). Considerable patron satisfaction with ID scanners in Geelong reflects public ‘apathy about having one’s driver’s license scanned’ (Holloman & Ponder 2007: 45) as identified in US literature. This also reveals immense trust that any personal data that is collected is unlikely to be ‘used outside of the scope of ensuring a safe and legal atmosphere within the establishments’ (Holloman & Ponder 2007: 45).
By contrast, privacy is a significant barrier to the dissemination of valuable information about trends in alcohol-related assault that reinforces the legitimacy of ID scanners as a crime prevention measure. The availability of data indicating where victims and perpetrators of assault had last consumed alcohol in the New South Wales city of Newcastle was considered vital to the development of targeted evidence-based interventions leading to a 30 percent reduction in alcohol-related assaults (Wiggers 2007; Wiggers et al. 2004). Such data could have additional value in measuring the displacement effects of any ID scanning or other forms of surveillance introduced into the night-time economy. 
They conclude that
Concepts of effectiveness based on deterrence, reduced anonymity and profiling have normalised digital ID verification as a legitimate form of social sorting. This study demonstrates that ID scanners are a tangible policy supported by what are currently poorly validated claims of ‘success’. The rapid introduction of these new technologies aimed at producing fast and discernible results has occurred with little consideration for the development of appropriate protocols regarding the collection, use, sharing, storage, maintenance, access to and destruction of digital information. Further, there has been limited oversight of how private venue operators or third-party security providers manage these databases independently of, and in conjunction with, the police.
They recommend establishment of a national working group to examine regulatory options for 'forms of population surveillance designed to prevent crime', with consideration of
  • A temporary moratorium on the use of ID scanners pending the development of an appropriate regulatory framework for data collection, storage, dissemination and privacy protection. 
  • The expansion of private security licensing to cover all personnel involved in using ID scanners, including the development of appropriate training and accountability measures. 
  • Specific policies and accountability processes for information sharing between private venue operators, security providers, police and criminal intelligence agencies. 
  • The provision of alternatives to ID scanning where patrons are unwilling to consent to the collection and storage of their personal information. 
  • The development of an independent and transparent complaints mechanism. 
  • The revision of current public order, summary offences, criminal, liquor licensing, administrative appeals and privacy regimes to develop an appropriate audit and compliance procedure. 
  • The promotion of ongoing local, state, national and international research into ID scanners, related electronic surveillance measures and their impact in preventing crime and alcohol-related harm.