No Identification Without Representation: Constraints on the Use of Biometric Identification Systems
May 9th, 2012 | By Innovya | Category: ArticlesThe tourist hoping to use her credit card in any part of the globe, the asylum seeker hoping to access social benefits in her host country, the banker hoping to move money from one stock market to another in real time—all have the same need. They must prove their identities and be certain of others’. Traditional means of proving identities are not dependable enough in most parts of the world and hence unfit for global digital networks. In this context, biometrics appears to offer a viable technological solution. However, the technology itself is subject to popular critique, warning of dystopian futures of overwhelming surveillance and loss of privacy.
By: Emilio Mordini, Andrew P. Rebera
Article first published online: 19 JAN 2012 DOI: 10.1111/j.1541-1338.2011.00535.x
© 2012 by The Policy Studies Organization
Abstract
The human species is again becoming nomadic. Each year, about two billion persons move across large geographic distances (not to mention people in “virtual mobility” through information and communication technology). Many of these people have weak or unreliable identification documents—and many poorer people in developing countries do not even have these documents. In 2000, the United Nations Children’s Fund (UNICEF) calculated that 50 million babies (41% of births worldwide) were not registered at birth. In this scenario, a personal identification scheme based on birth registration and state-issued passports is less and less tenable. Biometrics appears to offer a viable technological solution. However, the technology itself is subject to popular critique, warning of dystopian futures of overwhelming surveillance and loss of privacy. The best answer to those who fear an Orwellian future is to engage with the technology and seek to ensure that biometric identification systems are developed in positive ways. We suggest that identification schemes become problematic when the reciprocity of identification goes unnoticed, forgotten, or (what is worse) is intentionally bypassed. The dynamics of identification should be reciprocal, dialogical, and involving mutual recognition. In the traditional political domain, this is the recognition by the state of a citizen and by the citizen of the state. In the digital age, identification systems must increasingly transcend geopolitical borders. A globally recognized identification scheme is therefore a necessity. However, it is merely the nature of the borders that has changed here—not the nature of identification. Our call will be: no identification without representation.
1. Introduction
The development of methods for establishing, communicating, and authenticating the identity of individuals has been driven by a number of factors, not least the desire of governments to efficiently exercise power over their subjects. No doubt this was spurred by the development of concepts of the state, the individual, the citizen (and so forth). The modern state presupposes, in addition to technologies of identification, the “creation of a [. . .] people open to the scrutiny of officialdom” (Caplan & Torpey, 2001, p. 1).
That the need for reliable methods of identification arose at all is presumably connected with the birth of the first urban societies during the so-called “Neolithic Revolution” (Childe, 1950). At this point, the human species, hitherto nomadic, began to settle down (so to speak). Societies developed robust geographical roots, population densities increased, and social hierarchies developed. Growing societal complexity, alongside developments in artisanship and intrasocietal and intersocietal trade, would have made the identification of the trustworthy—as well as the detection of the untrustworthy—increasingly vital to the normal functioning of these early societies.
Of course it remains true that the major interventions of the state into the lives of its subjects—e.g., taxation, conscription, and the administration of authority/justice—need not proceed at the level of the individual. The necessity for individual identification is less severe when taxes, conscription, and so on are imposed collectively (Caplan & Torpey, 2001, p. 1). Specific individual identification has been more closely tied up with mass literacy and increased state bureaucracy (Goody, 1986). The Roman Empire, which was the first cosmopolitan society in the West, provided the first example of a universal identification system through a tripartite codified name scheme. In Europe during the Middle Ages, individuals were chiefly identified through passes and safe conducts issued by religious and civil authorities. Seals and handwriting mainly enforced the authenticity of these documents. The Modern era, which saw increased mobility associated with urbanization and industrialization, required more effective recognition schemes. These were administered by nation states. Indeed, the development of more robust identification schemes has to some extent progressed in parallel with the development of post-Westphalian polities, marking the passage from feudal to industrial society. Thus, as Valentin Groebner (2001, p. 16) notes: “by the beginning of the eighteenth century, the failure to carry [documents of origin and identity] while travelling was already an offense that could attract considerable penalties.”
After World War I, most European countries introduced systems of identity cards, incorporating facial photography (and, in some cases, also fingerprinting), as a tool for identifying people within their state borders. Yet even in countries where identity cards did not become mandatory, a new powerful driver for personal identification emerged: the need to identify and authenticate people entitled to receive social benefits. The welfare state, which first emerged in Northern Europe after World War II, is based on the provision of services via redistributionist taxation. Taxation and welfare provision both rely on robust and reliable systems of personal identification.
Following the agricultural, industrial, and welfare revolutions, we are now on the verge of a new epochal transition. The human species is again becoming nomadic. Each year, about two billion persons move across large geographic distances (not to mention people in “virtual mobility” through information and communication technology); approximately half cross international boundaries. The International Air Transport Association reported that their members carried 1.6 billion passengers in 2007, among which 699 million flew internationally (International Air Transportation Association, 2010). The United Nations World Tourism Organization (2009) estimated 924 million international tourist arrivals in 2008. International movements for permanent resettlement by immigrants, refugees, or asylum seekers, and temporary movement by migrant workers and others augment the total international movements each year. The International Labour Organization stated that in 2004, an estimated 175 million persons (3% of the world’s population) lived permanently outside their country of birth and that there were 81 million migrant workers (excluding refugees) globally (International Labour Organization, 2004).
Globalization has been characterized by the development of technologies dramatically transcending national control and regulation. This is not without consequence for traditional identification schemes. The globalized world is confronted with a huge mass of people with weak or absent identities. Most developing countries have weak or and unreliable documents—and many poorer people in these countries do not even have these documents. In 2000, UNICEF calculated that 50 million babies (41% of births worldwide) were not registered at birth. In Pakistan, Bangladesh, and Nepal, child registration at birth is not even mandatory. In this scenario, a personal identification scheme based on birth registration and state-issued passports is less and less tenable.
The tourist hoping to use her credit card in any part of the globe, the asylum seeker hoping to access social benefits in her host country, the banker hoping to move money from one stock market to another in real time—all have the same need. They must prove their identities and be certain of others’. Traditional means of proving identities are not dependable enough in most parts of the world and hence unfit for global digital networks. In this context, biometrics appears to offer a viable technological solution. However, the technology itself is subject to popular critique, warning of dystopian futures of overwhelming surveillance and loss of privacy.
For better or worse, biometric technologies are, it would seem, here to stay. The best answer to those who fear an Orwellian future is not merely to provide optimistic reassurances (which rarely reassure anyone) but rather to engage with the technology and seek to ensure that biometric identification systems are developed in positive ways. Our aim in this article is to articulate the positive potential of biometric identification schemes.
Orwellian fears are not unjustified, as we shall discuss. However, we will suggest that identification schemes become problematic when the reciprocity of identification goes unnoticed, forgotten, or (what is worse) is intentionally bypassed. The dynamics of identification should be reciprocal, dialogical, and involving mutual recognition. In the traditional political domain, this is the recognition by the state of a citizen and by the citizen of the state. In the digital age, identification systems must increasingly transcend geopolitical borders. A globally recognized identification scheme is therefore a necessity. However, it is merely the nature of the borders that has changed here, not the nature of identification. In this article, we will set out a framework for a system of identification. We begin in the following section by considering some objections to biometric identification systems. We will argue that where the objections are reasonable, this is not specifically attributable to the use of biometrics. We then (section 3) offer a partial defense of biometrics as a desirable candidate for a global identification system in the digital age. We do so by diagnosing the cause of the undesirable consequences associated with biometrics, discussed in section 2. Having identified the root problem, we conclude by proposing a guiding principle for the implementation of a global biometric identification system: no identification without representation.
2. Biometric Identification
Let us suppose that there is a group—relatively small in number—who would prefer that biometric systems of identification were not employed—that the technology had not been developed, or that other less intrusive forms of identification were more robustly suited to the digital age. For the sake of having a name, call this group “the anti-biometrics camp.” (We are speaking hypothetically here but it is not unreasonable to suppose that such a group exists.1) The anti-biometrics camp calls for the outlawing of biometric identification systems, or at least for such severe restrictions as to amount to a de facto outlawing of the technology. We will consider two arguments in the anti-biometrics camp’s arsenal: (i) that the technology is inherently demeaning and (ii) that the technology is a tool of the “surveillance society.”2
(i) Biometrics is Inherently Demeaning
“To identify” is a transitive verb. Correspondingly, the activity of identifying is relational: A identifies B, or, in the present case, Abiometrically identifies B. The question then, is what in this particular relational structure demeans B?
Standard means of gathering biometric data include: fingerprinting, facial scans, iris scans, voice recognition, hand vein recognition, gait analysis, keystroke pattern, and so on. Some of these means, it should be noted, are—if not inherently demeaning—at least conventionally associated with negative connotations. Fingerprinting is a clear example, being primarily associated with criminality (at least in some cultural areas). Furthermore, it is not unreasonable to suppose that most other biometrics—indeed the domain in general—are largely bound up in the popular imagination with criminality and crime detection. This is witnessed by the popularity of television shows centering on forensic pathology and criminology (e.g., CSI in the United States, or Silent Witness in the United Kingdom); while historically speaking, the associations trace back to the pioneers of anthropometrics such as Francis Galton (1822–1911) in the United Kingdom and Alphonse Bertillon (1853–1914) in France.3
In this respect, biometrics has something of an image problem. The problem is only deepened by high-profile examples in which biometrics plays a role—though it should be noted, for this is indicative of the extent of the image problem, that in many of these cases biometrics is incidental to the story.4 In a well-known case, Italian philosopher Georgio Agamben refused to enter the United States in protest at the United States Visitor and Immigrant Status Indicator (US-VISIT) program’s requirement for visitors to be fingerprinted and photographed (Agamben, 2008). In other cases too, U.S. immigration officials have been accused of heavy-handedness. In May 2011, 2000 elderly cruise passengers were fingerprinted and had retina scans during a 7-hour security check (Bloxham & Bingham, 2011); while Amir Khan, boxing’s IBF and WBA World Light Welterweight Champion, has complained of profiling and extensive security checks upon entering and re-entering the United States (Davies, 2011). Similarly, the use of biometrics was reported in connection with the recent controversy concerning the French government and the Roma community (Fraser, 2010). The biometrics connection was that the French authorities recorded biometric details of Roma migrants prior to repatriation. However, the primary newsworthy issue was not the use of biometrics but the alleged profiling of Roma communities, in apparent contravention of the European Union Charter of Fundamental Rights (BBC News, 2010a).
In only one of these cases (Agamben’s) is there any real suggestion that biometric technologies are themselves inherently demeaning. What is demeaning in the other cases is the way in which individuals are treated by the authorities. In the Agamben case, it is claimed that the gathering of biometric data is a form of bio-political tattooing, akin to the tattooing of Jews during the Holocaust. The implications of that association, if sound, are of course horrific. However, it is not entirely clear what Agamben’s argument really is. We will be brief as this is not the place (and even if it were, we have not the space) for detailed discussion. The difficulty is this. Agamben writes:
The problem [. . .] concerns the juridical-political status (it would be simpler, perhaps, to say bio-political) of citizens of the so-called democratic states where we live. [. . .] There has been an attempt the last few years to convince us to accept as the humane and normal dimensions of our existence, practices of control that had always been properly considered inhumane and exceptional. (2008, p. 201)
However, in and of itself, biometric identification is not obviously inhumane or exceptional. At root, biometric identification is nothing more than the identification of an individual by way of their physical or behavioral traits. This is something that we all do every day in normal interactions with our fellows. Here Agamben would probably argue that the body features that we use in everyday life to identify our fellows are biographical signs, embodied languages, “hot” media, which tell the biography of the subject. This is the case with human faces, body gestures, voices, odors, even wrinkles and scars, which are signatures of time on our skin. Biometrics are instead artificial signatures mechanically extracted from our bodies by impersonal technologies. They are “cool” media, which speak of our biology rather than our biography, so depersonalizing the subject. Biometrics—or so Agamben would argue—turn the persona into a bare body that becomessoma, as per the original meaning of this word in ancient Greek (roughly: corpse). However, Agamben would be wrong in two respects. First, the current trend in biometrics is to use more and more “hot” bodily features like face dynamics, gestures, behavioral traits, and so on. Second, the very notion of “bare body” is misleading. Human bodies are never pure bodies, they are always languages that tell stories: even the more remote physical features, even DNA, tell us much more than biological details, as is well illustrated by the huge amount of personal and medical information that can be elicited by any biometric signature (Mordini & Massari, 2008). Hence, it would seem that biometric identification is only inherently demeaning if normal interpersonal relationships are inherently demeaning—and while interpersonal relations can be demeaning, in the vast majority of cases, they are quite the opposite.
There is certainly some truth in what Agamben says, although we would suggest that he has mislocated it. As he says, this is a problem of juridical-political status—of bio-power relations; however, the core problem is not of biometrics. We can explore this by considering a response one might offer on behalf of the anti-biometrics camp. The analogy between identification in the sphere of interpersonal relationships and biometric identification is, the response goes, fundamentally flawed. True, normal interpersonal relationships are not demeaning, but then again one’s relationship to biometric identification devices—say, fingerprint scanners—is not interpersonal: fingerprint scanners are not persons, and our relations to them are not analogous to our relations to people. This is an important point with which we entirely agree.
Notice however that it does not support Agamben’s position. Agamben evokes Foucault, alluding to “a new bio-political era” (Agamben, 2008, p. 201). Foucault had articulated biopower as a “set of mechanisms through which the basic biological features of the human species became the object of a political strategy, of a general strategy of power” (Foucault, 2009, p. 1). However, he also pointed out that his analysis of the mechanisms of power involved, inter alia, investigating “between whom” (Foucault, 2009, p. 2) power is applied—that is to say (although Foucault may not have put it this way) that the power relations here are, in the final analysis, personal.5 So what is objectionable about U.S. immigration procedure is not, pace Agamben, the use of biometrics per se; rather, it is the way that they are deployed as an arbiter and medium of interpersonal relations.
(ii) Biometrics as a Tool of the “Surveillance Society”
Just as it would be wrong to suppose that biometric identification is in itself demeaning, so it would be wrong to suppose that biometric identification systems are inevitably geared toward surveillance. Or rather, it would be wrong to suppose that they are any more inevitably geared toward surveillance than are other forms of identification (any means of identification—biometrics, passports, proper names—can be used to keep track of an individual). However, the concerns regarding surveillance are perhaps more tangible than those discussed above, for the signs of surveillance are, in a physical sense, all around us.
The iconic technology of the surveillance society is probably the closed-circuit TV camera. Popular statistical folklore has it that there is one CCTV camera for every 14 British citizens and that the average Briton is caught on CCTV 300 times a day.6 In combination with CCTV, biometric identification technologies open the door to an enormous potential for surveillance. Facial recognition technology is the obvious candidate here, although recent work on identification by gait pattern has great potential too. As surveillance—spurred mainly by developments in biometrics—becomes increasingly automated, it inevitably becomes less and less focused. That is to say, we are not now dealing solely with the surveillance of antecedently identified suspects, but with the mass surveillance of society with the aim of identifying the suspicious among us. As an illustration of the dangers here, consider the use of search warrants. If authorities wish to search someone’s property, they need a warrant, which will, if all is working well, be issued only if sufficient evidentiary justification of reasonable suspicion can be produced. However, compare the surveillance of the 2001 Super Bowl. This measure identified, from a crowd of 100,000, just 19 individuals with criminal records. To have a criminal record is not a crime. And in a civilized country, to have a criminal record could not be considered sufficient evidentiary justification of reasonable suspicion that one might commit some (unspecified) crime. There is, it would seem, cause for concern here.
In the West, the United Kingdom and the United States have particularly poor records in this regard. In 2007, Privacy International rated the United Kingdom the worst performing European Union (EU) state in terms of privacy protection and surveillance, categorizing it alongside the United States, China, and Russia as “endemic surveillance societies” (Privacy International, 2007). The dangers have not gone unrecognized. In its response to the House of Commons’ Home Affairs Committee’s report, “A Surveillance Society” (Home Affairs Committee, 2008), the UK Government articulated a number of commitments to increase data minimization and protection and to ensure balance, responsibility, and transparency in the employment of surveillance technologies (Secretary of State for the Home Department, 2008). This was a welcome pronouncement in the light of Privacy International’s findings. (By 2010, the United Kingdom’s ranking had been downgraded from “endemic” to merely “extensive” surveillance [Privacy International, 2010]. This is better but still not good.)
Orwellian fears of all-pervasive surveillance are exacerbated by a number of factors. As mentioned above, developments in biometric technologies mean that identification can increasingly take place at a distance, via gait, for example (BBC News, 2010b). Second, biometrics and surveillance are increasingly commercial domains and, through public–private interaction, this can give rise to the impression that responsibility for good practice is passed from directly elected, and hence accountable, public officials, to businessmen and executives, whose primary goals are not best practice but profit, and whose responsibilities are to shareholders first, citizens second. (Note that this is not an objection to commercialization or the private sector in general; we merely note that different structures of accountability apply in the public and private sectors.) A further concern is with the speed with which new surveillance technologies are implemented—the worry being that insufficient consultation takes place. Examples include the case of the Visionics Corporation, which offered their facial recognition systems to the Tampa Police Department for a year free of charge “in an effort to build a market among municipalities,” prompting the American Civil Liberties Union of Florida to comment that “This is yet another example of technology outpacing the protection of people’s civil liberties” (Canedy, 2001). While in the United Kingdom, biometrics and CCTV in schools have been a cause of consternation. The Guardian, to give just one example, recently carried a story reporting various problematic cases of CCTV in schools (not only in classrooms but also in the bathrooms). A teachers’ union officer is reported to have commented: “There are lots of schools that install CCTV and don’t know the rules—and the companies who supply it don’t feel the need to tell them” (Harris, 2011).
Whatever may be the facts of the cases mentioned above, the fears are real. And the retort that if you have nothing to hide you have nothing to fear does not hold water. Privacy, secrecy, and freedom from surveillance have no necessary connection with shame or wrongdoing (Bok, 1989). Moreover, as the example of Brandon Mayfield (who was wrongly connected to the 2004 Madrid bombings on the basis of a false-positive fingerprint match) attests, it may well happen that you have nothing to hide but everything to fear; Mayfield was arrested, imprisoned, and claims to have been threatened with the death penalty. Eventually the U.S. government agreed to compensate him to the tune of $2 million (Eggen, 2006).
Fears of a surveillance society are not, then, unjustified. However, to reiterate, as with the objection in section (i), there is no necessaryconnection between biometrics and surveillance. The anti-biometrics camp is, we argue, misguided. Their position is vulnerable to two telling objections. First, in practical terms, their proposal is extremely unrealistic. The biometrics industry is well established: it has gained much support at governmental level (if not at public level), and although the technology is not flawless, it is advancing rapidly and appears to be the most likely candidate for a global identification system in the digital age. Second, the anti-biometrics camp misconstrues the conceptual link between biometrics and the potential undesirable consequences (individuals being demeaned and subject to surveillance) mentioned above. The source of the trouble is not, we will claim, biometric identification systems themselves but the way in which they are implemented.
3. Biometric Identification as a Reciprocal Relation
In this section, we make a partial defense of biometrics as a desirable candidate for a global identification system in the digital age, diagnosing the cause of the undesirable consequences associated with biometrics, as discussed above. Having identified the root problem, we will conclude by proposing a guiding principle for the implementation of a global identification system based on biometrics:no identification without representation.
Are biometric identification schemes desirable? We think they can be, if suitably regulated. As described above, biometrics can lead to bad outcomes. However, that does not seem to be a solid reason for opposing them. Consider the following analogy (or rather:disanalogy). The anti-biometrics camp associates biometrics with a number of undesirable consequences and argues that since the consequences are so undesirable, the benefits of biometrics do not outweigh the risks: therefore, biometrics should be outlawed. Compare the controversy in the United States over the Second Amendment right to bear arms.
“Guns don’t kill people, people kill people,” or so the bumper sticker goes. There is a clear—but limited—sense in which this is obviously true: guns do not normally kill people unless somebody pulls the trigger (and notice that we do not normally punish the gun). However, as the comedian Eddie Izzard notes, “If you just stood there and yelled BANG, I don’t think you’d kill too many people” (Izzard, Jordan, & Swanson, 2004). When somebody is shot, the gun obviously plays a rather important role: they would not have been shot without it. To oppose restrictions on gun ownership on the basis that guns do not kill people, people kill people, is plainly wrongheaded. Other things being equal, the fewer guns at large, the less likely one is to be a victim of gun crime. Indeed, it is trivially true that if there were no guns—if Mr. Gunpowder had never invented the stuff—there would be no gun crime. Hence, it would seem, abstracting from complexities, that there is a prima facie case to be made for banning guns.7 It is alas true that even if there were no gun crime, there would still be crime and people would still get murdered. However, if there are fewer murders—even just one fewer—that is definitely a good thing.
An evangelical biometrician might adorn her bumper with a sticker reading “biometrics don’t identify people, people identify people” (it is not so catchy as its cousin, but zealotry sees past such trifles). Will the Izzard response work here? It is true that if contemporary automated biometric identification systems did not exist, less identification would occur (for one of the claims of biometrics is that it provides a potential mode of identity management to those who currently have no identity documents at all). So in the absence of biometrics, there would, trivially, be no biometric identification and there would very likely be less identification overall. However, while lowering the murder rate is highly desirable, lowering the identification rate is not desirable. In the developed world, most of us do not want to be without identity management. We need secure identification systems in order to travel, to communicate and transact across the internet, and to claim welfare or tax-relief. We do not want no identification, we simply want a more reliable, less objectionable system than we have at present. And in the developing world, people with weak identity documents need robust alternatives. Thus, the point of this dis-analogy between guns and biometrics is that while shooting people is inherently undesirable, identifying people is not.8
If we accept that identification is desirable and that biometrics are neither inherently demeaning, nor any more geared toward surveillance than other identification systems, it behooves us to investigate biometric identification more carefully. What positives can it bring?
The advantages of biometric identification systems are a function of limitations of more traditional schemes. Mariana Muzzi (2010, p. 2) reports that “Around 51 million births go unregistered every year in developing countries, which translates to one in three children globally.” Recording the identity of these children is important: “Children whose births are not registered at birth are not able to claim the services and protection to which they have a right on a full and equal basis with children who are registered at birth” (Muzzi, 2010, p. 3). That is, one needs to establish one’s identity in order to claim and secure access to one’s rights. This is also the rhetoric behind India’s Unique Identification Number scheme (although what the reality is apart from the rhetoric is a matter of dispute—see, e.g., Ramakumar [2010]). Moreover, it is, to some extent, borne out by the transfer of welfare and entitlements direct to individuals in a large number of African counties (Devarajan & Giugale, 2011; Gelb & Decker, 2012)—such services being made cheaper and more secure by biometric means.
In such cases as these, we observe a reciprocal relation between identifier and identified. Schematically, we may put the point like this. “Aidentifies B,” but within this relational structure is the potential for reciprocity: A‘s identification of B can imply B‘s acknowledgement of A. This reciprocity is not essential. For example, if A is a geologist and B a fossil, the fossil does not acknowledge the geologist. However, where there is agency on the parts of both A and B, the relational structure “A identifies B” can imply reciprocity. Our claim is that “Aidentifies B” not only can imply reciprocity but should imply reciprocity. Let us try to be a little more specific. What exactly do we mean by “reciprocity?”
(i) Symmetry
By the reciprocity of identification we mean to imply a certain symmetry: if A identifies B, A should also be identified to B. This is a sort of exchange of information, but it grounds only a weak symmetry since no specifics as to the balance of information are implied. A may demand from B more—or at any rate different—information than A divulges. The rights and wrongs of the balance here will depend on circumstances.
(ii) Bestowal and Acknowledgment of Rights
In identifying B, A must bestow rights upon B, of which B is aware (or of which she can reasonably be expected to come to be aware).9 The system of registering children at birth is a fine example of this imperative. The child is registered and in that very act is documented as a bearer of rights. The provision of birth certificates may be thought of as the initial bestowal of rights. Subsequent identification transactions between A and B cannot, in general, be thought of as bestowing rights (although some may: e.g., A‘s issuing B a driving license could be thought of as A bestowing the legal right to drive—identifying B as a legal driver). However, these subsequent transactions may be thought of as A‘s acknowledging (explicitly or implicitly) the rights previously bestowed upon B (alternatively they might be thought of as premised upon A‘s earlier bestowal of rights upon B10).
(iii) Transparency
In normal circumstances, formal identification—i.e., identification by governments, commercial operations, and others—should be transparent. That is to say, it should be clear to B that they are being identified by A. The demands of transparency may depend on circumstances. For instance, if a border control agent asks for your passport, you do not normally ask them for proof of their identity and position—their identity and position are sufficiently clear in the context (they have the uniform, they are sitting in the booth at the airport, and so on). Or again, a store using CCTV to identify shoplifters need not be so transparent as to ask each customer to sign an informed consent form. However, they should, for example, display signs informing customers that they are being recorded. What level of transparency is appropriate to different contexts is a difficult issue which we have no intention of addressing here. We make the point in rather vague terms: identification should be tolerably transparent, given the circumstances.
Certainly, there is more to be said here. However, for present purposes, the above should suffice. In short, we argue that, at a minimum, the process of identification should: (i) be symmetrical; (ii) involve the bestowal upon, and acknowledgement of, the rights of the identified; and (iii) be transparent and open to the identified.
We conclude by looking at three problematic applications of biometric identification technologies. In each there is something objectionable. We will argue that what is objectionable is attributable to the failure to satisfy the reciprocity of identification.
Case 1: The Surveillance Society— Above, we identified a number of concerns regarding ubiquitous surveillance in a biometric future. The extent of CCTV coverage in public places is one cause for concern. However, the fact that you are often on film is not, in and of itself, any more objectionable than the fact that other people often observe you. Moreover, the fact that data regarding your whereabouts can be retrieved many years later is not, in and of itself, objectionable (if we were more observant and had better memories, the same data would be retrievable by testimony). Rather, what is objectionable about the profusion of CCTV cameras in public places is that, first, one does not know who is identifying one (although one knows that it is a formal identification—i.e., that it is a government or commercial agency, and so the situation is quite unlike being observed by a passerby in the street); second, it is not always apparent where the cameras are, so that one is not necessarily aware that one is under observation. These two problems are, respectively, failures to honor (i), the symmetry requirement, and (iii), the transparency requirement.
Similarly, the increased administration of surveillance by the private sector can be understood as objectionable in virtue of a failure to respect requirements (i) and (ii). Do the students in classrooms observed by cameras know which company is recording their images? If not, symmetry is not respected (or at least the symmetry is too weak given the context). Moreover, it will tend to be the case that if requirement (i) is not sufficiently well respected, requirement (ii) will not be either. Do private companies bestow any rights upon the students they identify? No.
To continue with the example of cameras in schools, one may object that the identification transactions do not involve the private companies. The private companies merely facilitate a transaction between the school and its students. If so, the use of CCTV in classrooms is, in some respects, less objectionable. Indeed, it is less objectionable to the extent that the situation is more in accord with requirements (i)–(iii).11
Finally, the concern that surveillance technologies are implemented too quickly, in advance of detailed consultation, may be understood as the concern that there is insufficient understanding of whether the administration of the new technology complies with requirements (i)–(iii). If the employment of new technology outstrips consultation as to its merits and demerits, the framework imposed by requirements (i)–(iii) cannot be guaranteed.
Case 2: The Afghan Villager— The New York Times reports that information on 1.5 million Afghans and 2.2 million Iraqis is now held in databases administered by U.S., NATO, and local forces (Shanker, 2011). In general, military–civilian identification transactions are liable to be problematic. This is evidenced by the prima facie double standards reported in the story: “While the systems [employed in Afghanistan and Iraq] are attractive to American law enforcement agencies, there is serious legal and political opposition to imposing routine collection on American citizens” (Shanker, 2011). We do not wish to comment on the specifics of the Afghan or Iraqi cases but speak, rather, about identification transactions between military forces and civilians in general, taking as inspiration the arresting image that accompanies the New York Times‘ article.
The photograph shows an aged, grey-bearded Afghan villager having his iris scanned. His right eye is shut, while the left is held open by the gloved hand of an American soldier. There are three soldiers in the picture: one holds the eye open, one holds the camera capturing the iris scan, and the other holds his hand above the villager’s forehead for shade. This shading hand is the only American flesh we see, all else is khaki, and we see no faces. It seems somehow significant that the uncovered hand casts a shadow not for the benefit of the villager, but in order that the iris camera is not “blinded” by the sunlight. It is a poignant image. What makes it so striking is the imbalance it portrays: they are soldiers, he is a civilian; they are three, he is one; his one eye is shut, the other held open.
Leaving aside the politics of the war in Afghanistan, as well as the particulars of the actual facts depicted in the photograph, this image—as a symbolic evocation—can only sound warning bells to those who are wary of biometrics. They will view it, not unreasonably, as a representation of the individual utterly overcome by faceless forces he is powerless to resist. Viewed in this way, what is objectionable about the image can be understood in terms of requirements (i)–(iii). First, this identification transaction is not symmetrical: the soldiers in the photo appear as anonymous.12 Second, the military cannot normally (in democratic societies at least) bestow rights upon individuals. Hence, requirement (ii) is not met.13 Third, it is entirely possible that military personnel operating overseas do not speak the local language. If so, it is not difficult to see how requirement (iii) might fail to be met. How can you explain to someone that they are being enrolled in a biometric database for whatever purpose, if you share no languages?14
The image of the Afghan villager is startling and, from a certain perspective, disturbing. What is disturbing about it is, we claim, that it is easy to read the story of the picture as if it involved the violation of the three requirements of the reciprocity of identification as described above.
Case 3: Bio-Political Tattooing— Recall that Agamben claims that the biometric enrollment required by the US-VISIT program is part of an attempt “to convince us to accept as the humane and normal dimensions of our existence, practices of control that had always been properly considered inhumane and exceptional” (Agamben, 2008, p. 201). We suggested that, in and of itself, biometric identification is neither inhumane nor exceptional. Rather, what is objectionable about U.S. immigration procedure is the manner in which it functions as an arbiter and medium of interpersonal relations. We are now in a position to make this a little clearer, with reference to requirements (i)–(iii).
Transparency does not appear to be an issue here. The procedures required as a condition of entry were sufficiently transparent that Agamben could consider them, judge them unreasonable, and decline to enter (although of course it is inconvenient not to enter a country having crossed the Atlantic to reach it). One could also make a case that symmetry is respected. The heart of the problem is requirement (ii), the bestowal and acknowledgment of rights.
Agamben is correct that there is something exceptional about the US-VISIT program. In broad terms, US-VISIT is simply an identification scheme: visitors must have their identities recorded upon entry. That is not exceptional: travel to almost any country and you will be identified as you cross the border. What is exceptional—what is not common to all border crossings—is that US-VISIT is not satisfied with the standard identity documents issued by other states but wishes to enrol individuals in a database wholly unconnected with their home state. This is an additional identification transaction, distinct from the presentation of a passport or visa. (We acknowledge of course that the United States is not the only state to request additional documentation or information to enter.) As such, if we are correct that identification should be governed by reciprocity as per requirements (i)–(iii), one ought to have some rights bestowed or acknowledged in the course of this additional identification transaction. But what rights does the entrant receive? She receives the privilege of entering the United States and enjoys the protection of the laws. However, that is something that, in pre-US-VISIT times, she would have received by virtue of the standard passport-based identification transaction; and (mutatis mutandis) it is something she still receives in most other countries by virtue of nothing more than a passport-based identification transaction. In this way, requirement (ii) is not met in the case of the additional biometric identification transaction that US-VISIT demands. However, pace Agamben, what is objectionableabout the additional biometric identification transaction is a failure of reciprocity—not anything specifically to do with biometric identification itself.
In insisting that the relational structure of identification be understood as reciprocal—i.e., that it involve symmetry adequate to the context; that it involve the bestowal upon, or acknowledgment of, rights of the identified individual; and that it be transparent—we are, in effect, insisting that identification be tightly bound to recognition of rights. If A may legitimately demand identifying data of B, A must acknowledge the rights of B. A owes something to B, which B may legitimately demand.
To speak in such terms is, of course, somewhat theoretical. In slightly more practical terms, our point can be made using the example of birth certificates. Registering births is a good thing because, as the child’s identity is registered, it becomes a bearer of rights; the child is, from the point of registration, hence, entitled to various protections by the state. However, as the statistics of birth registration attest, traditional means of identification are failing; and even where identification methods succeed, they are pressurized by the transition to the digital age. Biometric identification systems promise to be, from a technological point of view, adequate to the challenges of the digital age; and if they can be rolled out in accordance with the ideal of the reciprocity of identification, they will be a force for good. Rights will be protected because they will be enshrined in the very process of the state’s identification of its citizens. The state represents its citizens asbearers of rights. The power relations here may not be exactly equal, but they are at least reciprocal. Abuses of power arise where the reciprocity of identification falters.
Civil liberty advocates should not be frightened if it happens—as it has (Giroday, 2010)—that the head of Interpol calls for a globally verifiable electronic identity card (e-ID) system for migrant workers. This is a challenge and an opportunity because a global system of electronic identification is ethically and politically tenable only if it is sided by a global system for claiming fundamental rights and civil liberties; and this becomes increasingly feasible as we can provide everybody with secure and reliable identification. You can tax people, or alternatively identify them, only provided that you give them a political representation. In other words, identification is legitimate as far as it becomes instrumental to the enforcement of their political and civil rights. Identification without civil liberties would be no less abusive than taxation without representation. This is a lesson that—in a time of “tea party protesters”—it is worth remembering.15
- 1 We intend no reference to any actual group of that name, should such exist.
- 2 We do not discuss the other major concern, namely that the technology could be highly invasive of privacy and has the potential to reveal a huge amount of sensitive personal information regarding, say, health, background, lifestyle, and others. In this regard, there is some overlap with the issues discussed in section 2(ii).
- 3 See, e.g., Kaluszynski (2001), Joseph (2001), as well as Project Bertillon athttp://www.criminocorpus.cnrs.fr/bertillon/enter_uk.html.
- 4 In all cases bar the first of the following, biometrics is fairly obviously incidental. Even in the first (concerning Agamben) we feel that, ultimately, what is objectionable does not pertain to biometrics.
- 5 We are here using the word “personal” (and cognates) in a very wide sense, according to which relations between the state and an individual may be described as personal.
- 6 In the spirit of the (tongue in cheek) adage that 88.5% of statistics are made up, the veracity of these statistics is explored inAaronovitch (2009) and Channel 4 News (2008).
- 7 There are of course a variety of issues at play in this controversy. Our point is merely that there is a prima facie case to be answered. Whether it can be answered and how is not a point of interest here.
- 8 It is of course the case that if the primary goal of bearing arms is not to shoot people but, say, to shoot targets for sport, or to have something nice to hang above the mantelpiece, then we might construe the analogy differently. As mentioned in an earlier note, we are not really concerned with the rights and wrongs of gun ownership here.
- 9 How strongly one reads “bestow” here is likely to depend upon one’s answers to antecedent questions in political philosophy. A certain variety of social contract theorist might suppose that the citizen is, in some sense, born of identification by the state; in which case bestowal here is more or less literal: the state creates the citizen as a bearer of rights. Alternatively, if one holds that rights accrue to the individual independently of the state, identification by the state will involve less of a bestowal and more of an acknowledgment of rights. Our use of “bestow” is not intended to prejudge any of these foundational matters in political ontology and philosophy.
- 10 In the Kantian mode: the initial bestowal of rights serves as a condition of the possibility of all subsequent legitimate identification transactions between A and B, state and citizen.
- 11 The use of CCTV in classes would still be objectionable on the grounds of data minimization. What need is there for cameras in classes? If it is for discipline, or for teacher-training, these ends could be met by less intrusive methods. We will not explore the issues.
- 12 The soldiers appear in the photograph as anonymous. In reality they are (presumably) not, but have at least their names or numbers shown on their uniforms. Let us emphasize again that we are not discussing what or who is literally depicted in the photo, but its wider symbolic significance as an iconic image of the dark side of biometric identification transactions.
- 13 Again, the facts in Afghanistan may be different, but we are not discussing Afghanistan.
- 14 Once more: we are not discussing Afghanistan. Perhaps the actual soldiers in the photograph speak the villager’s language—perhaps he speaks theirs. Perhaps there has been a huge publicity campaign about the identification system.
- 15 This work has been funded by two European Commission research grants, RISE—Rising panEuropean and International Awareness of Biometrics and Security Ethics (GA230389), and TABULA RASA—Trusted Biometrics under Spoofing Attacks (GA 257289).
About the Authors
Emilio Mordini is the founding director of the Centre for Science, Society and Citizenship in Rome, Italy. He was Professor of Bioethics in the Medical School of the University of Rome “La Sapienza” (1994–2005), and a member (1994–2000) and secretary (2000–2004) of the Bioethical Commission of the Italian National Research Council. Since 1992 he has served as a contractor in quite a number of European Commission (EC) funded projects. His current board participations include: the Biometric Sector Federation of the Italian Confederation of Education and Knowledge Companies; the Committee for Standardization in ICT Focus Group on Biometrics; and the EC expert group on “ethical and regulatory challenges to science and research policy at the global level.” His research interests include ethical and social implications of security technology policies, and the ethics and policy of biometrics and emerging identification technology. His main publications include Ageing and Invisibility (IOS Press, 2010, edited with P. de Hert), and Second Generation Biometrics: the Ethical and Social Context (Springer, 2011, edited with D. Tzovaras).
Andrew P. Rebera (DPhil, University of Sussex) is a researcher at the Centre for Science, Society and Citizenship in Rome, Italy. His research interests include identity, privacy and data protection, surveillance and security, as well as philosophical issues in logic and metaphysics.
References
- Aaronovitch, D. (2009). The strange case of the surveillance cameras. The Times. Retrieved fromhttp://www.timesonline.co.uk/tol/comment/columnists/david_aaronovitch/article5834725.ece (accessed August 1, 2011; subscription required).
- Agamben, G. (2008). No to bio-political tattooing. Communication and Critical/Cultural Studies, 5(2), 201–202. Reproduced from Le Monde (2004, January 10).
- BBC News. (2010a). EU warns France of action over Roma. Retrieved from http://www.bbc.co.uk/news/world-europe-11437361(accessed August 1, 2011).
- BBC News. (2010b). Walk “could track down criminals,” researchers claim. Retrieved from http://www.bbc.co.uk/news/uk-england-hampshire-11025318 (accessed August 1, 2011).
- Bloxham, A., & Bingham, B. (2011). Cruise passengers tell of seven-hour security “revenge” nightmare. The Daily Telegraph. Retrieved from http://www.telegraph.co.uk/travel/travelnews/8559732/Cruise-passengers-tell-of-seven-hour-security-revenge-nightmare.html (accessed August 1, 2011).
- Bok, S. (1989). Secrets. New York: Vintage Books.
- Canedy, D. (2001). Tampa scans the faces in its crowds for criminals. New York Times. Retrieved fromhttp://www.nytimes.com/2001/07/04/us/tampa-scans-the-faces-in-its-crowds-for-criminals.html (accessed August 1, 2011).
- Caplan, J., & Torpey, J. (Eds.) (2001). Documenting individual identity. Princeton, NJ: Princeton University Press.
- Channel 4 News. (2008). FactCheck: How many CCTV cameras? Retrieved fromhttp://www.channel4.com/news/articles/society/factcheck+how+many+cctv+cameras/2291167.html (accessed August 1, 2011).
- Childe, V. G. (1950). The urban revolution. The Town Planning Review, 21, 3–17.
- Davies, G. A. (2011). Amir Khan furious at overzealous US immigration. The Daily Telegraph. Retrieved fromhttp://www.telegraph.co.uk/sport/othersports/boxingandmma/amir-khan/8566662/Amir-Khan-furious-at-overzealous-US-immigration.html (accessed August 1, 2011).
- Devarajan, S., & Giugale, M. (2011). How Africa can extract big benefits for everyone from natural resources. The Guardian. Retrieved from http://www.guardian.co.uk/global-development/poverty-matters/2011/jun/29/africa-extracting-benefits-from-natural-resources?INTCMP=SRCH (accessed August 2, 2011).
- Eggen, D. (2006). U.S. settles suit filed by Ore. lawyer. The Washington Post. Retrieved from http://www.washingtonpost.com/wp-dyn/content/article/2006/11/29/AR2006112901179.html (accessed August 2, 2011).
- Foucault, M. (2009). Security, territory, population: Lectures at the College de France 1977–1978. Translated by Graham Burchell. Basingstoke: Palgrave Macmillan.
- Fraser, C. (2010). France fingerprints Roma deportees to prevent return. Retrieved from http://www.bbc.co.uk/news/world-europe-11450831 (accessed August 1, 2011).
- Gelb, A., & Decker, C. (2012). Cash at your fingertips: Biometric technology for transfers in developing countries. Review of Policy Research, 29(1) 89–115.
- Giroday, S. B. (2010). ID WORLD Abu Dhabi 2010. Retrieved from: http://www.idworldabudhabi.com/index.php?id=records10ad(accessed August 23, 2011).
- Goody, J. (1986). The logic of writing and the organization of society. Cambridge: Cambridge University Press.
- Groebner, V. (2001). Describing the person, reading the signs in late medieval and Renaissance Europe: Identity papers, vested figures, and the limits of identification, 1400–1600. In J. Caplan & J. Torpey (Eds.), Documenting individual identity (pp. 15–27). Princeton, NJ: Princeton University Press.
- Harris, J. (2011). School surveillance: How big brother spies on pupils. The Guardian. Retrieved fromhttp://www.guardian.co.uk/uk/2011/jun/09/schools-surveillance-spying-on-pupils?INTCMP=SRCH (accessed August 1, 2011).
- Home Affairs Committee. (2008). A surveillance society? Retrieved fromhttp://www.publications.parliament.uk/pa/cm200708/cmselect/cmhaff/58/58ii.pdf (accessed August 1, 2011).
- International Air Transportation Association. (2010). Fact sheet: IATA—International air transport association. Retrieved fromhttp://www.iata.org/pressroom/facts_figures/fact_sheets/iata.htm (accessed August 18, 2011).
- International Labour Organization. (2004). Towards a fair deal for migrant workers in the global economy. International labour conference, 92nd session, report VI. Retrieved from http://www.ilo.org/global/publications/ilo-bookstore/order-online/books/WCMS_PUBL_9221130436_EN/lang-en/index.htm (accessed August 18, 2011).
- Izzard, E., Jordan, L., & Swanson, C. (Producers), Jordan, L. (Director). (2004). Dress to kill [DVD]. United Kingdom: Universal Pictures UK.
- Joseph, A. M. (2001). Anthropometry, the police expert, and the Deptford murders: The contested introduction of fingerprinting for the identification of criminals in Late Victorian and Edwardian Britain. In J. Caplan & J. Torpey (Eds.), Documenting individual identity(pp. 164–183). Princeton, NJ: Princeton University Press.
- Kaluszynski, M. (2001). Republican identity: Bertillonage as government technique. In J. Caplan & J. Torpey (Eds.), Documenting individual identity (pp. 123–138). Princeton, NJ: Princeton University Press.
- Mordini, E., & Massari, S. (2008). Body, biometrics and identity. Bioethics, 22, 488–498.
Direct Link:
- Muzzi, M. (2010). UNICEF good practices in integrating birth registration into health systems (2000–2009). Retrieved fromhttp://www.unicef.org/protection/Birth_Registration_Working_Paper(2).pdf (accessed August 2, 2011).
- Privacy International. (2007). Leading surveillance societies in the EU and the world 2007. Retrieved fromhttps://www.privacyinternational.org/article/leading-surveillance-societies-eu-and-world-2007 (accessed August 1, 2011).
- Privacy International. (2010). European privacy and human rights. Retrieved from https://www.privacyinternational.org/ephr (accessed August 1, 2011).
- Ramakumar, R. (2010). The unique ID project in India: A skeptical note. In A. Kumar & D. Zhang (Eds.), Ethics and policy of biometrics: ICEB 2010 (pp. 153–167). Berlin: Springer.
- Secretary of State for the Home Department. (2008). Government response to the home affairs committee: A surveillance society?Retrieved from http://www.official-documents.gov.uk/document/cm74/7449/7449.pdf (accessed August 1, 2011).
- Shanker, T. (2011). To track militants, U.S. has system that never forgets a face. New York Times. Retrieved fromhttp://www.nytimes.com/2011/07/14/world/asia/14identity.html?_r=2&scp=1&sq=biometrics&st=cse (accessed August 3, 2011).
- United Nations World Tourism Organization. (2009). UNWTO world tourism barometer, 2009. Retrieved fromhttp://unwto.org/facts/eng/pdf/barometer/UNWTO_Barom09_1_en_excerpt.pdf (accessed 18 August, 2011).
f
.jpg "Arthur I. Cyr")
